diff --git a/data/csv2_format_data.csv b/data/csv2_format_data.csv index ab32a867..54cc37eb 100644 --- a/data/csv2_format_data.csv +++ b/data/csv2_format_data.csv @@ -1,29 +1,29 @@ -INDEX:28:,ul_col:28:,dbl_col_2:26:,bool_col:6:,str_col:25:,xint_col:28:,dbl_col:27: -123450,123450,0.998,1,4% of something,1,1.2345 -123451,123451,0.3456,1,Description 4/5,2,2.2345 -123452,123452,0.056,1,This is bad,3,3.2345 -123450,123450,0.15678,0,3.4% of GDP,4,4.2345 -123455,123455,0.00345,0,Market drops,5,5.2345 -123450,123450,0.923,1,Market pulls back,3,3 -123449,123449,0.06743,,$15 increase,7,0.9999 -123450,123450,0.1,,Running fast,3,10 -123451,123451,0.0056,,C++14 development,9,4.25 -123450,123450,0.07865,,Some explanation,10,0.009 -123452,123452,0.0111,,More strings,3,8 -123450,123450,0.1002,,Bonds vs. Equities,2,2.2222 -123455,123455,-0.8888,,Almost done,3,3.3333 -123450,123450,0.14,,XXXX04,14,11 -123454,123454,0.0456,,XXXX2,2,5.25 -123450,123450,0.078654,,XXXX3,2,1.009 -123450,123450,-0.8999,,XXXX4,2,2.111 -123457,123457,0.8002,,XXXX4,3,9 -123458,123458,-0.9888,,XXXX5,2,3.2222 -123459,123459,0.2,,XXXX6,3,4.3333 -123450,123450,0.1056,,XXXX7,3,12 -123441,123441,0.87865,,XXXX10,3,6.25 -123442,123442,-0.6999,,XXXX11,3,2.009 -123432,123432,0.4111,,XXXX02,3,3.111 -123450,123450,0.1902,,XXXX03,36,10 -123450,123450,-0.4888,,,2,4.2222 -123435,123435,,,,45,5.3333 -123450,123450,,,,2, +INDEX:28:,ul_col:28:,dbl_col_2:26:,bool_col:6:,str_col:25:,xint_col:28:,dbl_col:27:,char_col:27:,uchar_col:27: +123450,123450,0.998,1,4% of something,1,1.2345,C,B +123451,123451,0.3456,1,Description 4/5,2,2.2345,23,B +123452,123452,0.056,1,This is bad,3,3.2345,^,& +123450,123450,0.15678,0,3.4% of GDP,4,4.2345,F,250 +123455,123455,0.00345,0,Market drops,5,5.2345,120,B +123450,123450,0.923,1,Market pulls back,3,3,78,B +123449,123449,0.06743,,$15 increase,7,0.9999,,B +123450,123450,0.1,,Running fast,3,10,C,B +123451,123451,0.0056,,C++14 development,9,4.25,C, +123450,123450,0.07865,,Some explanation,10,0.009,C,B +123452,123452,0.0111,,More strings,3,8,C,B +123450,123450,0.1002,,Bonds vs. Equities,2,2.2222,C,40 +123455,123455,-0.8888,,Almost done,3,3.3333,C,B +123450,123450,0.14,,XXXX04,14,11,C,B +123454,123454,0.0456,,XXXX2,2,5.25,C,B +123450,123450,0.078654,,XXXX3,2,1.009,, +123450,123450,-0.8999,,XXXX4,2,2.111,C, +123457,123457,0.8002,,XXXX4,3,9,C,B +123458,123458,-0.9888,,XXXX5,2,3.2222,,B +123459,123459,0.2,,XXXX6,3,4.3333,C,B +123450,123450,0.1056,,XXXX7,3,12,C,B +123441,123441,0.87865,,XXXX10,3,6.25,C,B +123442,123442,-0.6999,,XXXX11,3,2.009,C,B +123432,123432,0.4111,,XXXX02,3,3.111,C,B +123450,123450,0.1902,,XXXX03,36,10,C,B +123450,123450,-0.4888,,,2,4.2222,C,B +123435,123435,,,,45,5.3333,C,B +123450,123450,,,,2,,C,B diff --git a/data/sample_data.csv b/data/sample_data.csv index 2cfa271a..51cbe221 100644 --- a/data/sample_data.csv +++ b/data/sample_data.csv @@ -8,3 +8,9 @@ str_col:28::XXXX10,XXXX11,XXXX01,XXXX02,XXXX03,XXXX6,XXXX7,Running fast, dbl_col:28::2.009,3.111,10,4.2222,5.3333,12,6.25,10,0.9999,1.2345,4.2345,3,8,3.3333,2.2345,4.25,3.2345,0.009,1.111,5.25,11,5.2345,2.2222,1.009,2.111,9,3.2222,4.3333, dbl_col_2:28::0.87865,-0.6999,0.4111,0.1902,-0.4888,0.2,0.1056,0.1,0.06743,0.998,0.15678,0.923,0.0111,-0.8888,0.3456,0.0056,0.056,0.07865,-0.9999,0.0456,0.14,0.00345,0.1002,0.078654,-0.8999,0.01119,0.8002,-0.9888, bool_col:28::0,1,1,1,0,0,1,0,0,0,1,0,1,0,0,0,0,0,0,0,1,0,1,0,0,1,0,0, +Map 1:4::3{label one 1:123.0|label one 2:-782.5|label one 3:444.44},3{label two 1:123.0|label two 2:-782.5|label two 3:444.44},3{label three 1:123.0|label three 2:-782.5|label three 3:444.44},3{label four 1:123.0|label four 2:-782.5|label four 3:444.44}, +Unordered Map:4::3{Key one 1:123.0|Key one 2:-782.5|Key one 3:444.44},3{Key two 1:123.0|Key two 2:-782.5|Key two 3:444.44},3{Key three 1:123.0|Key three 2:-782.5|Key three 3:444.44},3{Key four 1:123.0|Key four 2:-782.5|Key four 3:444.44}, +Str Vec:4::4[bbb|aaa|zzz|ddd],4[aaa|bbb|ccc|www],4[123|abc|345|list],3[bbb|aaa|zzz], +Double Set:4::3[123.0|-782.5|444.44],3[1:123.0|-782.5|:444.44],3[123.0|-782.5|444.44],4[123.0|-782.5|444.44|100.5], +Str Set:4::3[123.0|-782.5|444.44],3[1:123.0|-782.5|:444.44],3[123.0|-782.5|444.44],4[123.0|-782.5|444.44|100.5], +Z Score:4::10[1.95474040557|0.552535091086|0.775388936446|-0.561817339812|0.106794118727|-0.153218675013|-0.896114748672|-1.72258101434|-0.301804546072|0.246077772077],10[-0.985180680575|-0.338649566179|1.37000434149|0.831246802651|-0.415610988193|1.06213106869|0.554158098662|0.507981245453|-1.55472278822|-1.03135753378],10[-1.39575784008|-1.62506351709|-0.907239380237|-0.159508137551|0.807541881212|0.937157205458|0.578236204203|0.717820563726|-0.0398797142361|1.0866927346],10[1.94246107491|-0.062340594565|0.246115232403|-1.24462409799|-0.190844664632|0.8115331407|0.381021476571|-1.60448155299|-0.422151990754|0.143311976349], diff --git a/data/sample_data.json b/data/sample_data.json index 6c4b0467..dd77b71f 100644 --- a/data/sample_data.json +++ b/data/sample_data.json @@ -6,5 +6,7 @@ "This is a test"]}, "col_2":{"N":12,"T":"double","D":[8,9,10,11,12,13,14,15,16,17,18, 777.78]}, "col_1":{"N":12,"T":"double","D":[1,2,3,4,5,6,7,8,9,10,11, - 55.55]} + 55.55]}, +"col_char":{"N":12,"T":"char","D":[8,C,F,$,,8,120,88,h,u,18, 78]}, +"col_uchar":{"N":12,"T":"uchar","D":[8,C,F,$,,8,220,88,h,&,18, 255]} } diff --git a/data/sample_data_2.json b/data/sample_data_2.json index e41a462d..7973f0b8 100644 --- a/data/sample_data_2.json +++ b/data/sample_data_2.json @@ -6,5 +6,7 @@ "This is a test"]}, "col_2_2":{"N":12,"T":"double","D":[8,9,10,11,12,13,14,15,16,17,18, 777.78]}, "col_1_2":{"N":12,"T":"double","D":[1,2,3,4,5,6,7,8,9,10,11, - 55.55]} + 55.55]}, +"col_char":{"N":12,"T":"char","D":[8,C,F,$,,8,120,88,h,u,18, 78]}, +"col_uchar":{"N":12,"T":"uchar","D":[8,C,F,$,,8,220,88,h,&,18, 255]} } diff --git a/data/sample_data_no_index.csv b/data/sample_data_no_index.csv index d3af792b..f75ff652 100644 --- a/data/sample_data_no_index.csv +++ b/data/sample_data_no_index.csv @@ -7,3 +7,5 @@ str_col_no_idx:28::XXXX10,XXXX11,XXXX01,XXXX02,XXXX03,XXXX6,XXXX7,Runnin dbl_col_no_idx:28::2.009,3.111,10,4.2222,5.3333,12,6.25,10,0.9999,1.2345,4.2345,3,8,3.3333,2.2345,4.25,3.2345,0.009,1.111,5.25,11,5.2345,2.2222,1.009,2.111,9,3.2222,4.3333, dbl_col_2_no_idx:28::0.87865,-0.6999,0.4111,0.1902,-0.4888,0.2,0.1056,0.1,0.06743,0.998,0.15678,0.923,0.0111,-0.8888,0.3456,0.0056,0.056,0.07865,-0.9999,0.0456,0.14,0.00345,0.1002,0.078654,-0.8999,0.01119,0.8002,-0.9888, bool_col_no_idx:28::0,1,1,1,0,0,1,0,0,0,1,0,1,0,0,0,0,0,0,0,1,0,1,0,0,1,0,0, +char_col:6::C,%,120,,65,! +uchar_col:6::C,%,250,,65,! diff --git a/data/sample_data_no_index.json b/data/sample_data_no_index.json index dd25c366..2a0aecfc 100644 --- a/data/sample_data_no_index.json +++ b/data/sample_data_no_index.json @@ -5,5 +5,7 @@ "This is a test"]}, "col_2_no_idx":{"N":12,"T":"double","D":[8,9,10,11,12,13,14,15,16,17,18, 777.78]}, "col_1_no_idx":{"N":12,"T":"double","D":[1,2,3,4,5,6,7,8,9,10,11, - 55.55]} + 55.55]}, +"col_char_no_idx":{"N":12,"T":"char","D":[8,C,F,$,,8,120,88,h,u,18, 78]}, +"col_uchar_no_idx":{"N":12,"T":"uchar","D":[8,C,F,$,,8,220,88,h,&,18, 255]} } diff --git a/data/sample_data_string_index.csv b/data/sample_data_string_index.csv index 67417729..d486a7ab 100644 --- a/data/sample_data_string_index.csv +++ b/data/sample_data_string_index.csv @@ -8,3 +8,12 @@ str_col:28::XXXX10,XXXX11,XXXX01,XXXX02,XXXX03,XXXX6,XXXX7,Running fast, dbl_col:28::2.009,3.111,10,4.2222,5.3333,12,6.25,10,0.9999,1.2345,4.2345,3,8,3.3333,2.2345,4.25,3.2345,0.009,1.111,5.25,11,5.2345,2.2222,1.009,2.111,9,3.2222,4.3333, dbl_col_2:28::0.87865,-0.6999,0.4111,0.1902,-0.4888,0.2,0.1056,0.1,0.06743,0.998,0.15678,0.923,0.0111,-0.8888,0.3456,0.0056,0.056,0.07865,-0.9999,0.0456,0.14,0.00345,0.1002,0.078654,-0.8999,0.01119,0.8002,-0.9888, bool_col:28::0,1,1,1,0,0,1,0,0,0,1,0,1,0,0,0,0,0,0,0,1,0,1,0,0,1,0,0, +char_col:6::C,%,120,,65,!, +uchar_col:6::C,%,250,,65,!, +Map 1:4::3{label one 1:123.0|label one 2:-782.5|label one 3:444.44},3{label two 1:123.0|label two 2:-782.5|label two 3:444.44},3{label three 1:123.0|label three 2:-782.5|label three 3:444.44},3{label four 1:123.0|label four 2:-782.5|label four 3:444.44}, +Unordered Map:4::3{Key one 1:123.0|Key one 2:-782.5|Key one 3:444.44},3{Key two 1:123.0|Key two 2:-782.5|Key two 3:444.44},3{Key three 1:123.0|Key three 2:-782.5|Key three 3:444.44},3{Key four 1:123.0|Key four 2:-782.5|Key four 3:444.44}, +Str Vec:4::4[bbb|aaa|zzz|ddd],4[aaa|bbb|ccc|www],4[123|abc|345|list],3[bbb|aaa|zzz], +Double Set:4::3[123.0|-782.5|444.44],3[1:123.0|-782.5|:444.44],3[123.0|-782.5|444.44],4[123.0|-782.5|444.44|100.5], +Str Set:4::3[123.0|-782.5|444.44],3[1:123.0|-782.5|:444.44],3[123.0|-782.5|444.44],4[123.0|-782.5|444.44|100.5], +Z Score:4::10[1.95474040557|0.552535091086|0.775388936446|-0.561817339812|0.106794118727|-0.153218675013|-0.896114748672|-1.72258101434|-0.301804546072|0.246077772077],10[-0.985180680575|-0.338649566179|1.37000434149|0.831246802651|-0.415610988193|1.06213106869|0.554158098662|0.507981245453|-1.55472278822|-1.03135753378],10[-1.39575784008|-1.62506351709|-0.907239380237|-0.159508137551|0.807541881212|0.937157205458|0.578236204203|0.717820563726|-0.0398797142361|1.0866927346],10[1.94246107491|-0.062340594565|0.246115232403|-1.24462409799|-0.190844664632|0.8115331407|0.381021476571|-1.60448155299|-0.422151990754|0.143311976349], + diff --git a/docs/HTML/DataFrame.html b/docs/HTML/DataFrame.html index cf9c5152..d124ca2a 100644 --- a/docs/HTML/DataFrame.html +++ b/docs/HTML/DataFrame.html @@ -88,7 +88,7 @@
  • To start off on basic operations, see Hello World
  • Also, see DataFrame Library Types
  • DataFrame has both sync and async interfaces, latter returning C++ std::futures
    • -
  • Read views, visitors, multithreading, and memory alignment sections below, before getting serious about this library
    • +
  • Read multithreading, views, visitors, and memory alignment sections below, before getting serious about this library

    • @@ -181,6 +181,10 @@

    API Reference with code samples

    bucketize_async( ) + + clear( ) + + col_name_to_idx( ) @@ -265,6 +269,10 @@

    API Reference with code samples

    get_data_by_idx( 2 ) + + get_data_by_like( 2 ) + + get_data_by_loc( 2 ) @@ -309,6 +317,10 @@

    API Reference with code samples

    get_view_by_idx( 2 ) + + get_view_by_like( 2 ) + + get_view_by_loc( 2 ) @@ -349,6 +361,10 @@

    API Reference with code samples

    has_column( 2 ) + + inversion_count( ) + + is_equal( ) @@ -509,6 +525,10 @@

    API Reference with code samples

    sort_async( 5 ) + + swap( ) + + to_string( ) diff --git a/docs/HTML/get_data_by_like.html b/docs/HTML/get_data_by_like.html new file mode 100644 index 00000000..cecbd60d --- /dev/null +++ b/docs/HTML/get_data_by_like.html @@ -0,0 +1,297 @@ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
    Signature Description Parameters
    + 
    
+template<StringOnly T, typename ... Ts>
+DataFrame
+get_data_by_like(const char *name,
+                 const char *pattern,
+                 bool case_insensitive = false,
+                 char esc_char = '\\') const;
+
    +     		+ This method does a basic Glob-like pattern matching (also similar to SQL like clause) to filter data in the named column. It returns a new DataFrame. Each element of the named column is checked against a Glob-like matching logic

    + + Globbing rules:
    +  '*' Matches any sequence of zero or more characters.
    +  '?' Matches exactly one character.
    +  [...] Matches one character from the enclosed list of characters.
    +  [^...] Matches one character not in the enclosed list.

    + + With the [...] and [^...] matching, a ']' character can be included in the list by making it the first character after '[' or '^'. A range of characters can be specified using '-'. Example: "[a-z]" matches any single lower-case letter. To match a '-', make it the last character in the list.

    + + Hint: To match '*' or '?', put them in "[]". Like this: abc[*]xyz matches "abc*xyz" only
    + NOTE: This could be, in some cases, n-squared. But it is pretty fast with moderately sized strings. I have not tested this with huge/massive strings.
    +     		+ T: Type of the named column. Based on the concept, it can only be either of these types: std::string, VirtualString, const char *, char *
    + Ts: List all the types of all data columns. A type should be specified in the list only once.
    + name: Name of the data column
    + pattern: Glob like pattern to use for matching strings
    + case_insensitive: If true, matching logic ignores case
    + esc_char: Character used for escape
    +
    + 
    
+template<StringOnly T, typename ... Ts>
+PtrView
+get_view_by_like(const char *name,
+                 const char *pattern,
+                 bool case_insensitive = false,
+                 char esc_char = '\\');
+
    +     		+ This is identical with above get_data_by_like(), but:
    +
      +
    1. The result is a view
      2. +
    2. Since the result is a view, you cannot call make_consistent() on the result.
      3. +
    + NOTE: There are certain operations that you cannot do with a view. For example, you cannot add/delete columns, etc.
    +     		+ T: Type of the named column. Based on the concept, it can only be either of these types: std::string, VirtualString, const char *, char *
    + Ts: List all the types of all data columns. A type should be specified in the list only once.
    + name: Name of the data column
    + pattern: Glob like pattern to use for matching strings
    + case_insensitive: If true, matching logic ignores case
    + esc_char: Character used for escape
    +
    + 
    
+template<StringOnly T, typename ... Ts>
+ConstPtrView
+get_view_by_like(const char *name,
+                 const char *pattern,
+                 bool case_insensitive = false,
+                 char esc_char = '\\') const;
+
    +     		+ Same as above view, but it returns a const view. You can not change data in const views. But if the data is changed in the original DataFrame or through another view, it is refelcted in the const view. + + T: Type of the named column. Based on the concept, it can only be either of these types: std::string, VirtualString, const char *, char *
    + Ts: List all the types of all data columns. A type should be specified in the list only once.
    + name: Name of the data column
    + pattern: Glob like pattern to use for matching strings
    + case_insensitive: If true, matching logic ignores case
    + esc_char: Character used for escape
    +
    + 
    
+template<StringOnly T, typename ... Ts>
+DataFrame
+get_data_by_like(const char *name1,
+                 const char *name2,
+                 const char *pattern1,
+                 const char *pattern2,
+                 bool case_insensitive = false,
+                 char esc_char = '\\') const;
+
    +     		+ This does the same function as above get_data_by_like() but operating on two columns.
    +     		+ T: Type of both named columns. Based on the concept, it can only be either of these types: std::string, VirtualString, const char *, char *
    + Ts: List all the types of all data columns. A type should be specified in the list only once.
    + name1: Name of the first data column
    + name2: Name of the second data column
    + pattern1: Glob like pattern to use for matching strings for the first column
    + pattern2: Glob like pattern to use for matching strings for the second column
    + case_insensitive: If true, matching logic ignores case
    + esc_char: Character used for escape
    +
    + 
    
+template<StringOnly T, typename ... Ts>
+PtrView
+get_view_by_like(const char *name1,
+                 const char *name2,
+                 const char *pattern1,
+                 const char *pattern2,
+                 bool case_insensitive = false,
+                 char esc_char = '\\');
+
    +     		+ This is identical with above get_data_by_like(), but:
    +
      +
    1. The result is a view
      2. +
    2. Since the result is a view, you cannot call make_consistent() on the result.
      3. +
    + NOTE: There are certain operations that you cannot do with a view. For example, you cannot add/delete columns, etc.
    +     		+ T: Type of both named columns. Based on the concept, it can only be either of these types: std::string, VirtualString, const char *, char *
    + Ts: List all the types of all data columns. A type should be specified in the list only once.
    + name1: Name of the first data column
    + name2: Name of the second data column
    + pattern1: Glob like pattern to use for matching strings for the first column
    + pattern2: Glob like pattern to use for matching strings for the second column
    + case_insensitive: If true, matching logic ignores case
    + esc_char: Character used for escape
    +
    + 
    
+template<StringOnly T, typename ... Ts>
+ConstPtrView
+get_view_by_like(const char *name1,
+                 const char *name2,
+                 const char *pattern1,
+                 const char *pattern2,
+                 bool case_insensitive = false,
+                 char esc_char = '\\') const;
+
    +     		+ Same as above view, but it returns a const view. You can not change data in const views. But if the data is changed in the original DataFrame or through another view, it is refelcted in the const view. + + T: Type of both named columns. Based on the concept, it can only be either of these types: std::string, VirtualString, const char *, char *
    + Ts: List all the types of all data columns. A type should be specified in the list only once.
    + name1: Name of the first data column
    + name2: Name of the second data column
    + pattern1: Glob like pattern to use for matching strings for the first column
    + pattern2: Glob like pattern to use for matching strings for the second column
    + case_insensitive: If true, matching logic ignores case
    + esc_char: Character used for escape
    +
    + +
    static void test_get_data_by_like()  {
+
+    std::cout << "\nTesting get_data_by_like( ) ..." << std::endl;
+
+    StlVecType<unsigned long>  idxvec = { 1UL, 2UL, 3UL, 10UL, 5UL, 7UL, 8UL, 12UL, 9UL, 12UL, 10UL, 13UL, 10UL, 15UL, 14UL };
+    StlVecType<std::string>    strvec1 =
+        { "345&%$abcM", "!@#$0987^HGTtiff\"", "ABFDTiy", "345&%$abcM", "!@#$0987^HGTtiff\"", "!@#$0987^HGTtiff\"", "345&%$abcM",
+          "!@#$0987^HGTtiff\"", "ABFDTiy", "345&%$abcM", "!@#$0987^HGTtiff\"", "ABFDTiy", "345&%$abcM", "ABFDTiy", "ABFDTiy" };
+    StlVecType<std::string>    strvec2 =
+        { "ABFDTiy", "!@#$0987^HGTtiff\"", "ABFDTiy", "345&%$abcM", "!@#$0987^HGTtiff\"", "ABFDTiy", "!@#$0987^HGTtiff\"",
+          "!@#$0987^HGTtiff\"", "ABFDTiy", "345&%$abcM", "!@#$0987^HGTtiff\"", "ABFDTiy", "345&%$abcM", "ABFDTiy", "ABFDTiy" };
+    StlVecType<int>            intvec = { 1, 2, 3, 10, 5, 7, 8, 12, 9, 12, 10, 13, 10, 15, 14 };
+    MyDataFrame                df;
+
+    df.load_data(std::move(idxvec),
+                 std::make_pair("str column 1", strvec1),
+                 std::make_pair("str column 2", strvec2),
+                 std::make_pair("int column", intvec));
+
+    auto    df_like2 =
+        df.get_data_by_like<std::string, std::string, int>("str column 1",
+                                                           "str column 2",
+                                                           "?*[0-9][0-9][0-9][0-9]?*",
+                                                           "?*[0-9][0-9][0-9][0-9]?*");
+
+    assert(df_like2.get_index().size() == 4);
+    assert(df_like2.get_index()[2] == 12);
+    assert(df_like2.get_column<int>("int column")[2] == 12);
+    assert(df_like2.get_column<std::string>("str column 1").size() == 4);
+    assert(df_like2.get_column<std::string>("str column 2").size() == 4);
+    assert((df_like2.get_column<std::string>("str column 1")[0] == "!@#$0987^HGTtiff\""));
+    assert((df_like2.get_column<std::string>("str column 1")[2] == "!@#$0987^HGTtiff\""));
+    assert((df_like2.get_column<std::string>("str column 2")[0] == "!@#$0987^HGTtiff\""));
+    assert((df_like2.get_column<std::string>("str column 2")[2] == "!@#$0987^HGTtiff\""));
+
+    auto    dfv_like2 =
+        df.get_view_by_like<std::string, std::string, int>("str column 1",
+                                                           "str column 2",
+                                                           "?*[0-9][0-9][0-9][0-9]?*",
+                                                           "?*[0-9][0-9][0-9][0-9]?*");
+
+    assert(dfv_like2.get_index().size() == 4);
+    assert(dfv_like2.get_index()[2] == 12);
+    assert(dfv_like2.get_column<int>("int column")[2] == 12);
+    assert(dfv_like2.get_column<std::string>("str column 1").size() == 4);
+    assert(dfv_like2.get_column<std::string>("str column 2").size() == 4);
+    assert((dfv_like2.get_column<std::string>("str column 1")[0] == "!@#$0987^HGTtiff\""));
+    assert((dfv_like2.get_column<std::string>("str column 1")[2] == "!@#$0987^HGTtiff\""));
+    assert((dfv_like2.get_column<std::string>("str column 2")[0] == "!@#$0987^HGTtiff\""));
+    assert((dfv_like2.get_column<std::string>("str column 2")[2] == "!@#$0987^HGTtiff\""));
+
+    dfv_like2.get_column<std::string>("str column 2")[3] = "ABC";
+    assert(dfv_like2.get_column<std::string>("str column 2")[3] == "ABC");
+    assert(df.get_column<std::string>("str column 2")[10] == "ABC");
+
+    auto    df_like1 =
+        df.get_data_by_like<std::string, std::string, int>("str column 1", "?*&%?*");
+
+    assert(df_like1.get_index().size() == 5);
+    assert(df_like1.get_index()[2] == 8);
+    assert(df_like1.get_column<int>("int column")[2] == 8);
+    assert(df_like1.get_column<std::string>("str column 1").size() == 5);
+    assert(df_like1.get_column<std::string>("str column 2").size() == 5);
+    assert((df_like1.get_column<std::string>("str column 1")[0] == "345&%$abcM"));
+    assert((df_like1.get_column<std::string>("str column 1")[2] == "345&%$abcM"));
+    assert((df_like1.get_column<std::string>("str column 2")[0] == "ABFDTiy"));
+    assert((df_like1.get_column<std::string>("str column 2")[2] == "!@#$0987^HGTtiff\""));
+}
+
    + + C++ DataFrame + + + + + diff --git a/docs/HTML/inversion_count.html b/docs/HTML/inversion_count.html new file mode 100644 index 00000000..4a2b5f9d --- /dev/null +++ b/docs/HTML/inversion_count.html @@ -0,0 +1,115 @@ + + + + + + + + + + + + + + + +
    Signature Description Parameters
    + 
    
+template<typename T, typename C = std::less<T>>
+size_type
+inversion_count(const char *col_name) const;
+
    +     		+ This retunrs the number of inversions in the named column.
    + For example, in a column that is already sorted, the number of inversions is zero. In a column that is sorted in reverse, the number of inversions is n(n - 1) / 2.
    +     		+ T: Data type of the named column
    + C: Type of the comparison functor defaulted to std::less
    + col_name: Name of the column
    +
    + +
    static void test_inversion_count()  {
+
+    std::cout << "\nTesting inversion_count(  ) ..." << std::endl;
+
+    using IntDataFrame = StdDataFrame<int>;
+
+    std::vector<int>    idx = { 0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17 };
+    std::vector<int>    i1 = { 0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17 };
+    std::vector<int>    i2 = { 17,16,15,14,13,12,11,10,9,8,7,6,5,4,3,2,1,0 };
+    std::vector<int>    i3 = { 1,0,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17 };
+    std::vector<int>    i4 = { 0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,17,16 };
+    std::vector<int>    i5 = { 1,0,2,3,4,5,6,7,8,9,10,11,12,13,14,15,17,16 };
+    std::vector<int>    i6 = { 17,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,0 };
+    std::vector<int>    i7 = { 0,1,2,3,4,5,6,10,8,9,7,11,12,13,14,15,16,17 };
+    std::vector<int>    i8 = { 0,1,2,15,4,5,6,7,8,9,10,11,12,13,14,3,16,17 };
+    std::vector<int>    i9 = { 2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2 };
+    std::vector<int>    i10 = { 2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,3 };
+    std::vector<int>    i11 = { 2,2,2,2,3,2,2,2,2,4,2,2,2,5,2,2,2,6 };
+    IntDataFrame        df;
+
+    df.load_data(std::move(idx),
+                 std::make_pair("i1", i1),
+                 std::make_pair("i2", i2),
+                 std::make_pair("i3", i3),
+                 std::make_pair("i4", i4),
+                 std::make_pair("i5", i5),
+                 std::make_pair("i6", i6),
+                 std::make_pair("i7", i7),
+                 std::make_pair("i8", i8),
+                 std::make_pair("i9", i9),
+                 std::make_pair("i10", i10),
+                 std::make_pair("i11", i11));
+
+    assert(df.inversion_count<int>("i1") == 0);
+    assert(df.inversion_count<int>("i2") == 153);
+    assert(df.inversion_count<int>("i3") == 1);
+    assert(df.inversion_count<int>("i4") == 1);
+    assert(df.inversion_count<int>("i5") == 2);
+    assert(df.inversion_count<int>("i6") == 33);
+    assert(df.inversion_count<int>("i7") == 5);
+    assert(df.inversion_count<int>("i8") == 23);
+    assert(df.inversion_count<int>("i9") == 153);
+    assert(df.inversion_count<int>("i10") == 136);
+    assert(df.inversion_count<int>("i11") == 110);
+
+    assert((df.inversion_count<int, std::greater<int>>("i1") == 153));
+    assert((df.inversion_count<int, std::greater<int>>("i2") == 0));
+}
    + + C++ DataFrame + + + + + diff --git a/docs/HTML/read.html b/docs/HTML/read.html index cc806e23..78cde5ee 100644 --- a/docs/HTML/read.html +++ b/docs/HTML/read.html @@ -44,6 +44,12 @@ It inputs the contents of a text file/stream into itself (i.e. DataFrame). Currently 3 formats (i.e. csv, csv2, json) are supported. See io_format documentation page
    +
    +  NOTE: If the DataFrame that is reading the file already has existing data columns, the file data will be added to the existing DataFrame columns.
+        If the file has a data column with the same name and type as a column in the DataFrame, the file data will replace the existing data column in the DataFrame.
+        If the file has a data column with the same name but different type as a column in the DataFrame, the behavior is undefined.
+        Obviously, if the DataFrame is empty none of these matters.
+

    CSV file format must be:
       INDEX:<Number of data points>:<Comma delimited list of values>
@@ -85,24 +91,27 @@
         
    
         In all formats the following data types are supported:
         
    -          float
-          double
-          longdouble  -- long double
-          int
-          uint        -- unsigned int
-          long
-          longlong    -- long long int
-          ulong       -- unsigned long
-          ulonglong   -- unsigned long long int
-          string
-          bool
-          DateTime    -- DateTime data in format of <Epoch seconds>.<nanoseconds> (1516179600.874123908)
+          float      -- float
+          double     -- double
+          longdouble -- long double
+          short      -- short int
+          ushort     -- unsigned short int
+          int        -- int
+          uint       -- unsigned int
+          long       -- long int
+          longlong   -- long long int
+          ulong      -- unsigned long int
+          ulonglong  -- unsigned long long int
+          char       -- char
+          uchar      -- unsigned char
+          string     -- std::string
+          string     -- const char *
+          string     -- char *
+          bool       -- bool
+          DateTime   -- DateTime data in format of <Epoch seconds>.<nanoseconds> (1516179600.874123908)
         
    
-        In case of io_format::csv2 the following additional types are also supported:
+        In case of io_format::csv2 and io_format::csv the following additional types are also supported:
         
    -          DateTimeAME    -- DateTime string printed in American style (MM/DD/YYYY HH:MM:SS.mmm)
-          DateTimeEUR    -- DateTime string printed in European style (YYYY/MM/DD HH:MM:SS.mmm)
-          DateTimeISO    -- DateTime string printed in ISO style (YYYY-MM-DD HH:MM:SS.mmm)
           dbl_vec        -- A vector of double precision values, The vector is printed as "s[d1|d2|...]"
                             where s is the size of the vector and d's are the double values.
           str_vec        -- A vector of std::string values, The vector is printed as "s[str1|str2|...]"
@@ -117,6 +126,13 @@
                             where s is the size of the map and k's and v's are keys and values.
         
    
 
+        In case of io_format::csv2 the following additional types are also supported:
+        
    +          DateTimeAME -- DateTime string printed in American style (MM/DD/YYYY HH:MM:SS.mmm)
+          DateTimeEUR -- DateTime string printed in European style (YYYY/MM/DD HH:MM:SS.mmm)
+          DateTimeISO -- DateTime string printed in ISO style (YYYY-MM-DD HH:MM:SS.mmm)
+        
    
+
         NOTE:: This version of read() can be substantially faster, especially for larger files, than if you open the file yourself and use the read() version below.
       
       
diff --git a/docs/HTML/remove_column.html b/docs/HTML/remove_column.html
index f36f170c..a75118f0 100644
--- a/docs/HTML/remove_column.html
+++ b/docs/HTML/remove_column.html
@@ -33,16 +33,17 @@
     
        
         
    
+template<typename T>
 void
 remove_column(const char *name);
         
    
       
       
-        It removes a column named name.
    
-        The actual data vector is not deleted, but the column is dropped from DataFrame
+        It removes the named column and frees up the memory space.
    
       
       
-        name: Column name
+        T: Type of the named column
    
+        name: Column name
    
       
     
 
@@ -50,56 +51,179 @@
     
        
         
    
+template<typename T>
 void
 remove_column(std::size_t index);
         
    
       
       
-        It removes a column based on index.
    
-        The actual data vector is not deleted, but the column is dropped from DataFrame
+        It removes a column based on index and frees up the memory space.
    
       
       
-        index: Column index
+        T: Type of the named column
    
+        index: Column index
    
       
     
 
-  
-
-  
    static void test_remove_column()  {
-
-    std::cout << "\nTesting remove_column() ..." << std::endl;
-
-    std::vector<unsigned long>  idx = { 123450, 123451, 123452, 123450, 123455, 123450, 123449 };
-    std::vector<double> d1 = { 1, 2, 3, 4, 5, 6, 7 };
-    std::vector<double> d2 = { 8, 9, 10, 11, 12, 13, 14 };
-    std::vector<double> d3 = { 15, 16, 17, 18, 19, 20, 21 };
-    std::vector<int>    i1 = { 22, 23, 24, 25 };
-    std::vector<std::string> s1 = { "11", "22", "33", "xx", "yy", "gg", "string" };
-    MyDataFrame         df;
-
-    df.load_data(std::move(idx),
-                 std::make_pair("col_1", d1),
-                 std::make_pair("col_2", d2),
-                 std::make_pair("col_3", d3),
-                 std::make_pair("col_int", i1),
-                 std::make_pair("col_str", s1));
+    
+    
+       
+        
    
+void
+clear();
+        
    
+      
+      
+        This removes all the index and data columns but doesn't necessarily free memeory space of underlying containers. After this call DataFrame will be empty.
    
+        It is very similar to std::vector clear()
    
+      
+      
+      
+    
 
-    df.write<std::ostream, double, int, std::string>(std::cout);
-    df.remove_column("col_2");
-    std::cout << "After removing column `col_2`" << std::endl;
-    df.write<std::ostream, double, int, std::string>(std::cout);
-    df.remove_column("col_str");
-    std::cout << "After removing column `col_str`" << std::endl;
-    df.write<std::ostream, double, int, std::string>(std::cout);
+    
+    
+       
+        
    
+void
+swap(DataFrame &other);
+        
    
+      
+      
+        This swaps all self's index and data columns with the ones in other
    
+        It is very similar to std::vector swap()
    
+      
+      
+        other: Another DataFrme of the same type
    
+      
+    
 
-    std::vector<double> d22 = { 8, 9, 10, 11, 12, 13, 14 };
+  
 
-    df.load_column("col_2", std::move(d22));
-    std::cout << "After adding back column `col_2`" << std::endl;
-    df.write<std::ostream, double, int, std::string>(std::cout);
-}
-
    
-  
+
    static void test_remove_column()  {
+
+    std::cout << "\nTesting remove_column() ..." << std::endl;
+
+    StlVecType<unsigned long>  idx =
+        { 123450, 123451, 123452, 123450, 123455, 123450, 123449 };
+    StlVecType<double>         d1 = { 1, 2, 3, 4, 5, 6, 7 };
+    StlVecType<double>         d2 = { 8, 9, 10, 11, 12, 13, 14 };
+    StlVecType<double>         d3 = { 15, 16, 17, 18, 19, 20, 21 };
+    StlVecType<int>            i1 = { 22, 23, 24, 25 };
+    StlVecType<std::string>    s1 = { "11", "22", "33", "xx", "yy", "gg", "string" };
+    MyDataFrame                df;
+
+    df.load_data(std::move(idx),
+                 std::make_pair("col_1", d1),
+                 std::make_pair("col_2", d2),
+                 std::make_pair("col_3", d3),
+                 std::make_pair("col_int", i1),
+                 std::make_pair("col_str", s1));
+
+    df.write<std::ostream, double, int, std::string>(std::cout);
+    df.remove_column<double>("col_2");
+    std::cout << "After removing column `col_2`" << std::endl;
+    df.write<std::ostream, double, int, std::string>(std::cout);
+    df.remove_column<std::string>("col_str");
+    std::cout << "After removing column `col_str`" << std::endl;
+    df.write<std::ostream, double, int, std::string>(std::cout);
+
+    StlVecType<double> d22 = { 8, 9, 10, 11, 12, 13, 14 };
+
+    df.load_column<double>("col_2", std::move(d22));
+    std::cout << "After adding back column `col_2`" << std::endl;
+    df.write<std::ostream, double, int, std::string>(std::cout);
+}
+
    
+
    +// -----------------------------------------------------------------------------
+
+static void test_clear()  {
+
+    std::cout << "\nTesting clear( ) ..." << std::endl;
+
+    StlVecType<unsigned long>  idx =
+        { 123450, 123451, 123452, 123453, 123454, 123455, 123456, 123457, 123458, 123459, 123460, 123461, 123462, 123466 };
+    StlVecType<double> d1 = { 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14 };
+    StlVecType<double> d2 = { 8, 9, 10, 11, 12, 13, 14, 20, 22, 23, 30, 31, 32, 1.89 };
+    StlVecType<double> d3 = { 15, 16, 17, 18, 19, 20, 21, 0.34, 1.56, 0.34, 2.3, 0.1, 0.89, 0.45 };
+    StlVecType<int>    i1 = { 22, 23, 24, 25, 99, 100, 101, 3, 2 };
+    StlVecType<std::string>    strvec = { "zz", "bb", "cc", "ww", "ee", "ff", "gg", "hh", "ii", "jj", "kk", "ll", "mm", "nn" };
+    MyDataFrame        df1;
+
+    df1.load_data(std::move(idx),
+                  std::make_pair("col_1", d1),
+                  std::make_pair("col_2", d2),
+                  std::make_pair("col_3", d3),
+                  std::make_pair("col_4", i1),
+                  std::make_pair("str_col", strvec));
+
+    StlVecType<unsigned long>  idx2 =
+        { 123450, 123451, 123452, 123453, 123454, 123455, 123456, 123457, 123458, 123459, 123460, 123461, 123462, 123466 };
+    StlVecType<double> d12 = { 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14 };
+    StlVecType<double> d22 = { 8, 9, 10, 11, 12, 13, 14, 20, 22, 23, 30, 31, 32, 1.89 };
+    StlVecType<double> d32 = { 15, 16, 17, 18, 19, 20, 21, 0.34, 1.56, 0.34, 2.3, 0.1, 0.89, 0.45 };
+    StlVecType<int>    i12 = { 22, 23, 24, 25, 99, 100, 101, 3, 2 };
+    StlVecType<std::string>    strvec2 = { "zz", "bb", "cc", "ww", "ee", "ff", "gg", "hh", "ii", "jj", "kk", "ll", "mm", "nn" };
+    MyDataFrame        df2;
+
+    df2.load_data(std::move(idx2),
+                  std::make_pair("col_1", d12),
+                  std::make_pair("col_2", d22),
+                  std::make_pair("col_3", d32),
+                  std::make_pair("col_4", i12),
+                  std::make_pair("str_col", strvec2));
+
+    df1.clear();
+    assert(df1.empty());
+    assert(df1.shapeless());
+    assert(df2.get_index()[4] == 123454);
+    assert(df2.get_column<int>("col_4")[7] == 3);
+    assert(df2.get_column<std::string>("str_col")[5] == "ff");
+
+    df1 = df2;
+    assert(df1.get_index()[4] == 123454);
+    assert(df1.get_column<int>("col_4")[7] == 3);
+    assert(df1.get_column<std::string>("str_col")[5] == "ff");
+}
+
    
+
    // -----------------------------------------------------------------------------
+
+static void test_swap()  {
+
+    std::cout << "\nTesting swap( ) ..." << std::endl;
+
+    StlVecType<unsigned long>  idx =
+        { 123450, 123451, 123452, 123453, 123454, 123455, 123456, 123457, 123458, 123459, 123460, 123461, 123462, 123466 };
+    StlVecType<double> d1 = { 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14 };
+    StlVecType<double> d2 = { 8, 9, 10, 11, 12, 13, 14, 20, 22, 23, 30, 31, 32, 1.89 };
+    StlVecType<double> d3 = { 15, 16, 17, 18, 19, 20, 21, 0.34, 1.56, 0.34, 2.3, 0.1, 0.89, 0.45 };
+    StlVecType<int>    i1 = { 22, 23, 24, 25, 99, 100, 101, 3, 2 };
+    StlVecType<std::string>    strvec = { "zz", "bb", "cc", "ww", "ee", "ff", "gg", "hh", "ii", "jj", "kk", "ll", "mm", "nn" };
+    MyDataFrame        df1;
+    MyDataFrame        df2;
+
+    df1.load_data(std::move(idx),
+                  std::make_pair("col_1", d1),
+                  std::make_pair("col_2", d2),
+                  std::make_pair("col_3", d3),
+                  std::make_pair("col_4", i1),
+                  std::make_pair("str_col", strvec));
+
+    assert(df2.empty());
+    assert(df2.shapeless());
+    assert(df1.get_index()[4] == 123454);
+    assert(df1.get_column<int>("col_4")[7] == 3);
+    assert(df1.get_column<std::string>("str_col")[5] == "ff");
+
+    df1.swap(df2);
+    assert(df1.empty());
+    assert(df1.shapeless());
+    assert(df2.get_index()[4] == 123454);
+    assert(df2.get_column<int>("col_4")[7] == 3);
+    assert(df2.get_column<std::string>("str_col")[5] == "ff");
+}
+
    
 
   C++ DataFrame
diff --git a/docs/HTML/write.html b/docs/HTML/write.html
index 86f607ef..143f31e6 100644
--- a/docs/HTML/write.html
+++ b/docs/HTML/write.html
@@ -44,7 +44,7 @@
    - It outputs the content of DataFrame into the stream o. Currently 3 formats (i.e. csv, csv2, json) are supported specified by the iof parameter.
    + It outputs the content of DataFrame into the stream o. Currently 3 formats (i.e. csv, csv2, json) are supported specified by the iof parameter.


    The CSV file format is written:
       INDEX:<Number of data points>:<Comma delimited list of values>
@@ -86,24 +86,27 @@
         
    
         In all formats the following data types are supported:
         
    -          float
-          double
-          longdouble  -- long double
-          int
-          uint        -- unsigned int
-          long
-          longlong    -- long long int
-          ulong       -- unsigned long
-          ulonglong   -- unsigned long long int
-          string
-          bool
-          DateTime    -- DateTime data in format of <Epoch seconds>.<nanoseconds> (1516179600.874123908)
+          float      -- float
+          double     -- double
+          longdouble -- long double
+          short      -- short int
+          ushort     -- unsigned short int
+          int        -- int
+          uint       -- unsigned int
+          long       -- long int
+          longlong   -- long long int
+          ulong      -- unsigned long int
+          ulonglong  -- unsigned long long int
+          char       -- char
+          uchar      -- unsigned char
+          string     -- std::string
+          string     -- const char *
+          string     -- char *
+          bool       -- bool
+          DateTime   -- DateTime data in format of <Epoch seconds>.<nanoseconds> (1516179600.874123908)
         
    
-        In case of io_format::csv2 the following additional types are also supported:
+        In case of io_format::csv2 and io_format::csv the following additional types are also supported:
         
    -          DateTimeAME    -- DateTime string printed in American style (MM/DD/YYYY HH:MM:SS.mmm)
-          DateTimeEUR    -- DateTime string printed in European style (YYYY/MM/DD HH:MM:SS.mmm)
-          DateTimeISO    -- DateTime string printed in ISO style (YYYY-MM-DD HH:MM:SS.mmm)
           dbl_vec        -- A vector of double precision values, The vector is printed as "s[d1|d2|...]"
                             where s is the size of the vector and d's are the double values.
           str_vec        -- A vector of std::string values, The vector is printed as "s[str1|str2|...]"
@@ -117,6 +120,14 @@
           str_dbl_unomap -- An unordered map of string keys to double precision values, The map is printed as "s{k1:v1|k2:v2|...}"
                             where s is the size of the map and k's and v's are keys and values.
         
    
+
+        In case of io_format::csv2 the following additional types are also supported:
+        
    +          DateTimeAME -- DateTime string printed in American style (MM/DD/YYYY HH:MM:SS.mmm)
+          DateTimeEUR -- DateTime string printed in European style (YYYY/MM/DD HH:MM:SS.mmm)
+          DateTimeISO -- DateTime string printed in ISO style (YYYY-MM-DD HH:MM:SS.mmm)
+        
    
+
       
       
         
    diff --git a/include/DataFrame/DataFrame.h b/include/DataFrame/DataFrame.h
index cbee0995..ef623a15 100644
--- a/include/DataFrame/DataFrame.h
+++ b/include/DataFrame/DataFrame.h
@@ -77,7 +77,7 @@ class   DataFrame : public ThreadGranularity {
 
 public:  // Construction
 
-    static constexpr std::size_t    align_value { std::size_t(H::align_value) };
+    static constexpr std::size_t   align_value { std::size_t(H::align_value) };
 
     template
     using AllocatorType = typename allocator_declare::type;
@@ -150,15 +150,42 @@ class   DataFrame : public ThreadGranularity {
     create_column(const char *name, bool do_lock = true);
 
     // It removes a column named name.
-    // The actual data vector is not deleted, but the column is dropped from
-    // DataFrame
     //
+    // T:
+    //   Type of the named column
+    // name:
+    //   Name of the column
+    //
+    template
     void
     remove_column(const char *name);
 
+    // T:
+    //   Type of the indexed column
+    // index:
+    //   Index of the column
+    //
+    template
     void
     remove_column(size_type index);
 
+    // This removes all the index and data columns but doesn't necessarily
+    // free memeory space of underlying containers. After this call DataFrame
+    // will be empty.
+    // It is very similar to std::vector clear()
+    //
+    void
+    clear();
+
+    // This swaps all self's index and data columns with the ones in other
+    // It is very similar to std::vector swap()
+    //
+    // other:
+    //   Another DataFrme of the same type
+    //
+    void
+    swap(DataFrame &other);
+
     // It renames column named from to to. If column from does not exist,
     // it throws an exception
     //
@@ -1598,7 +1625,8 @@ class   DataFrame : public ThreadGranularity {
     [[nodiscard]] DataFrame
     shift(size_type periods, shift_policy sp) const;
 
-    // This copies the named column into another vector and shifts it up or down
+    // This copies the named column into another vector and shifts it up
+    // or down
     // and returns it.
     // It is handy to create columns of shifted data in the dataframe for
     // machine-learning analysis
@@ -1987,9 +2015,6 @@ class   DataFrame : public ThreadGranularity {
     //   2) Since the result is a view, you cannot call make_consistent() on
     //      the result.
     //
-    // NOTE: Although this is a const method, it returns a view. So, the data
-    //       could still be modified through the returned view
-    //
     // T:
     //   Type of the named column
     // F:
@@ -2470,6 +2495,130 @@ class   DataFrame : public ThreadGranularity {
     [[nodiscard]] DataFrame
     get_data_by_sel(F &sel_functor, FilterCols&&... filter_cols) const;
 
+    // This method does a basic Glob-like pattern matching (also similar to
+    // SQL like clause) to filter data in the named column.
+    // It returns a new DataFrame. Each element of the named column is checked
+    // against a Glob-like matching logic
+    //
+    // Globbing rules:
+    //
+    //      '*'       Matches any sequence of zero or more characters.
+    //
+    //      '?'       Matches exactly one character.
+    //
+    //     [...]      Matches one character from the enclosed list of
+    //                characters.
+    //
+    //     [^...]     Matches one character not in the enclosed list.
+    //
+    // With the [...] and [^...] matching, a ']' character can be included
+    // in the list by making it the first character after '[' or '^'.  A
+    // range of characters can be specified using '-'.  Example:
+    // "[a-z]" matches any single lower-case letter. To match a '-', make
+    // it the last character in the list.
+    //
+    // Hints: to match '*' or '?', put them in "[]". Like this:
+    //        abc[*]xyz matches "abc*xyz" only
+    //
+    // NOTE: This could be, in some cases, n-squared. But it is pretty fast
+    //       with moderately sized strings. I have not tested this with
+    //       huge/massive strings.
+    //
+    // T:
+    //   Type of the named column. Based on the concept, it can only be either
+    //   of these types: std::string, VirtualString, const char *, char *
+    // Ts:
+    //   List all the types of all data columns. A type should be specified in
+    //   the list only once.
+    // name:
+    //   Name of the data column
+    // pattern:
+    //   Glob like pattern to use for matching strings
+    // case_insensitive:
+    //   If true, matching logic ignores case
+    // esc_char:
+    //   Character used for escape
+    //
+    template
+    [[nodiscard]] DataFrame
+    get_data_by_like(const char *name,
+                     const char *pattern,
+                     bool case_insensitive = false,
+                     char esc_char = '\\') const;
+
+    // This is identical with above get_data_by_like(), but:
+    //   1) The result is a view
+    //   2) Since the result is a view, you cannot call make_consistent() on
+    //      the result.
+    //
+    template
+    [[nodiscard]] PtrView
+    get_view_by_like(const char *name,
+                     const char *pattern,
+                     bool case_insensitive = false,
+                     char esc_char = '\\');
+
+    template
+    [[nodiscard]] ConstPtrView
+    get_view_by_like(const char *name,
+                     const char *pattern,
+                     bool case_insensitive = false,
+                     char esc_char = '\\') const;
+
+    // This does the same function as above get_data_by_like() but operating
+    // on two columns.
+    //
+    // T:
+    //   Type of both named columns. Based on the concept, it can only be
+    //   either of these types: std::string, VirtualString, const char *, char *
+    // Ts:
+    //   List all the types of all data columns. A type should be specified in
+    //   the list only once.
+    // name1:
+    //   Name of the first data column
+    // name2:
+    //   Name of the second data column
+    // pattern1:
+    //   Glob like pattern to use for matching strings for the first column
+    // pattern2:
+    //   Glob like pattern to use for matching strings for the second column
+    // case_insensitive:
+    //   If true, matching logic ignores case
+    // esc_char:
+    //   Character used for escape
+    //
+    template
+    [[nodiscard]] DataFrame
+    get_data_by_like(const char *name1,
+                     const char *name2,
+                     const char *pattern1,
+                     const char *pattern2,
+                     bool case_insensitive = false,
+                     char esc_char = '\\') const;
+
+    // This is identical with above get_data_by_like(), but:
+    //   1) The result is a view
+    //   2) Since the result is a view, you cannot call make_consistent() on
+    //      the result.
+    //
+    template
+    [[nodiscard]] PtrView
+    get_view_by_like(const char *name1,
+                     const char *name2,
+                     const char *pattern1,
+                     const char *pattern2,
+                     bool case_insensitive = false,
+                     char esc_char = '\\');
+
+    template
+    [[nodiscard]] ConstPtrView
+    get_view_by_like(const char *name1,
+                     const char *name2,
+                     const char *pattern1,
+                     const char *pattern2,
+                     bool case_insensitive = false,
+                     char esc_char = '\\') const;
+
     // It returns a DataFrame (including the index and data columns)
     // containing the data from uniform random selection.
     // random_policy determines the behavior of method.
@@ -2665,6 +2814,21 @@ class   DataFrame : public ThreadGranularity {
     [[nodiscard]] StringStats
     get_str_col_stats(const char *col_name) const;
 
+    // This retunrs the number of inversions in the named column. For example,
+    // in a column that is already sorted, the number of inversions is zero.
+    // In a column that is sorted in reverse, the number of inversions is
+    // n(n - 1) / 2.
+    //
+    // T:
+    //   Data type of the named column
+    // C:
+    //   Type of the comparison functor defaulted to std::less
+    // col_name:
+    //   Name of the column
+    //
+    template>
+    size_type inversion_count(const char *col_name) const;
+
     // This function returns a DataFrame indexed by std::string that provides
     // a few statistics about the columns of the calling DataFrame.
     // The statistics are:
diff --git a/include/DataFrame/Internals/DataFrame.tcc b/include/DataFrame/Internals/DataFrame.tcc
index 16680517..cc5d186a 100644
--- a/include/DataFrame/Internals/DataFrame.tcc
+++ b/include/DataFrame/Internals/DataFrame.tcc
@@ -182,222 +182,6 @@ DataFrame::shuffle(const StlVecType &col_names,
 
 // ----------------------------------------------------------------------------
 
-template
-template
-void DataFrame::
-fill_missing_value_(ColumnVecType &vec,
-                    const T &value,
-                    int limit,
-                    size_type col_num)  {
-
-    const size_type vec_size = vec.size();
-    int             count = 0;
-
-    if (limit < 0)
-        vec.reserve(col_num);
-    for (size_type i = 0; i < col_num; ++i)  {
-        if (limit >= 0 && count >= limit)  break;
-        if (i >= vec_size)  {
-            vec.push_back(value);
-            count += 1;
-        }
-        else if (is_nan(vec[i]))  {
-            vec[i] = value;
-            count += 1;
-        }
-    }
-    return;
-}
-
-// ----------------------------------------------------------------------------
-
-template
-template
-void DataFrame::
-fill_missing_ffill_(ColumnVecType &vec, int limit, size_type col_num)  {
-
-    const size_type vec_size = vec.size();
-
-    if (vec_size == 0)  return;
-
-    int count = 0;
-    T   last_value = vec[0];
-
-    for (size_type i = 1; i < col_num; ++i)  {
-        if (limit >= 0 && count >= limit)  break;
-        if (i >= vec_size)  {
-            if (! is_nan(last_value))  {
-                vec.reserve(col_num);
-                vec.push_back(last_value);
-                count += 1;
-            }
-            else  break;
-        }
-        else  {
-            if (! is_nan(vec[i]))
-                last_value = vec[i];
-            else if (! is_nan(last_value))  {
-                vec[i] = last_value;
-                count += 1;
-            }
-        }
-    }
-    return;
-}
-
-// ----------------------------------------------------------------------------
-
-template
-template::value ||
-                                 ! supports_arithmetic::value>::type*>
-void DataFrame::
-fill_missing_midpoint_(ColumnVecType &, int, size_type)  {
-
-    throw NotFeasible("fill_missing_midpoint_(): ERROR: Mid-point filling is "
-                      "not feasible on non-arithmetic types");
-}
-
-// ----------------------------------------------------------------------------
-
-template
-template::value &&
-                                 supports_arithmetic::value>::type*>
-void DataFrame::
-fill_missing_midpoint_(ColumnVecType &vec, int limit, size_type)  {
-
-    const size_type vec_size = vec.size();
-
-    if (vec_size < 3) [[unlikely]]  return;
-
-    int count = 0;
-    T   last_value = vec[0];
-
-    for (size_type i = 1; i < vec_size - 1; ++i)  {
-        if (limit >= 0 && count >= limit)  break;
-
-        if (! is_nan(vec[i]))
-            last_value = vec[i];
-        else if (! is_nan(last_value))  {
-            for (size_type j = i + 1; j < vec_size; ++j)  {
-                if (! is_nan(vec[j]))  {
-                    vec[i] = (last_value + vec[j]) / T(2);
-                    last_value = vec[i];
-                    count += 1;
-                    break;
-                }
-            }
-        }
-    }
-    return;
-}
-
-// ----------------------------------------------------------------------------
-
-template
-template
-void DataFrame::
-fill_missing_bfill_(ColumnVecType &vec, int limit)  {
-
-    const long  vec_size = static_cast(vec.size());
-
-    if (vec_size == 0)  return;
-
-    int count = 0;
-    T   last_value = vec[vec_size - 1];
-
-    for (long i = vec_size - 1; i >= 0; --i)  {
-        if (limit >= 0 && count >= limit)  break;
-        if (! is_nan(vec[i]))  last_value = vec[i];
-        if (is_nan(vec[i]) && ! is_nan(last_value))  {
-            vec[i] = last_value;
-            count += 1;
-        }
-    }
-    return;
-}
-
-// ----------------------------------------------------------------------------
-
-template
-template::value ||
-                                 ! supports_arithmetic::value>::type*>
-void DataFrame::
-fill_missing_linter_(ColumnVecType &, const IndexVecType &, int)  {
-
-    throw NotFeasible("fill_missing_linter_(): ERROR: Interpolation is "
-                      "not feasible on non-arithmetic types");
-}
-
-// ----------------------------------------------------------------------------
-
-template
-template::value &&
-                                 supports_arithmetic::value>::type*>
-void DataFrame::
-fill_missing_linter_(ColumnVecType &vec,
-                     const IndexVecType &index,
-                     int limit)  {
-
-    const long  vec_size = static_cast(vec.size());
-
-    if (vec_size < 3)  return;
-
-    int             count = 0;
-    const T         *y1 = &(vec[0]);
-    const T         *y2 = &(vec[2]);
-    const IndexType *x = &(index[1]);
-    const IndexType *x1 = &(index[0]);
-    const IndexType *x2 = &(index[2]);
-
-    for (long i = 1; i < vec_size - 1; ++i)  {
-        if (limit >= 0 && count >= limit)  break;
-        if (is_nan(vec[i]))  {
-            if (is_nan(*y2))  {
-                bool    found = false;
-
-                for (long j = i + 1; j < vec_size; ++j)  {
-                    if (! is_nan(vec[j]))  {
-                        y2 = &(vec[j]);
-                        x2 = &(index[j]);
-                        found = true;
-                        break;
-                    }
-                }
-                if (! found)  break;
-            }
-            if (is_nan(*y1))  {
-                for (long j = i - 1; j >= 0; --j)  {
-                    if (! is_nan(vec[j]))  {
-                        y1 = &(vec[j]);
-                        x1 = &(index[j]);
-                        break;
-                    }
-                }
-            }
-            vec[i] =
-                *y1 +
-                (static_cast(*x - *x1) / static_cast(*x2 - *x1)) *
-                (*y2 - *y1);
-            count += 1;
-        }
-        if (i < (vec_size - 2))  {
-            y1 = &(vec[i]);
-            y2 = &(vec[i + 2]);
-            x = &(index[i + 1]);
-            x1 = &(index[i]);
-            x2 = &(index[i + 2]);
-        }
-    }
-
-    return;
-}
-
-// ----------------------------------------------------------------------------
-
 template
 template
 void DataFrame::
@@ -503,40 +287,6 @@ void DataFrame::fill_missing (const DF &rhs)  {
 
 // ----------------------------------------------------------------------------
 
-template
-template
-void DataFrame::
-drop_missing_rows_(T &vec,
-                   const DropRowMap &missing_row_map,
-                   drop_policy policy,
-                   size_type threshold,
-                   size_type col_num)  {
-
-    size_type   erase_count = 0;
-    auto        dropper =
-        [&vec, &erase_count](const auto &iter) -> void  {
-            vec.erase(vec.begin() + (iter.first - erase_count));
-            erase_count += 1;
-        };
-
-    if (policy == drop_policy::all)  {
-        for (const auto &iter : missing_row_map)
-            if (iter.second == col_num)  dropper(iter);
-    }
-    else if (policy == drop_policy::any)  {
-        for (const auto &iter : missing_row_map)
-            if (iter.second > 0)  dropper(iter);
-    }
-    else if (policy == drop_policy::threshold)  {
-        for (const auto &iter : missing_row_map)
-            if (iter.second > threshold)  dropper(iter);
-    }
-
-    return;
-}
-
-// ----------------------------------------------------------------------------
-
 template
 template
 void DataFrame::
diff --git a/include/DataFrame/Internals/DataFrame_functors.h b/include/DataFrame/Internals/DataFrame_functors.h
index 53488d77..34d89eb4 100644
--- a/include/DataFrame/Internals/DataFrame_functors.h
+++ b/include/DataFrame/Internals/DataFrame_functors.h
@@ -52,7 +52,22 @@ struct consistent_functor_ : DataVec::template visitor_base  {
 template
 struct shrink_to_fit_functor_ : DataVec::template visitor_base  {
 
-    inline shrink_to_fit_functor_ ()  {   }
+    inline shrink_to_fit_functor_()  {   }
+
+    template
+    void operator() (T &vec) const;
+};
+
+// ----------------------------------------------------------------------------
+
+template
+struct remove_column_functor_ : DataVec::template visitor_base  {
+
+    inline remove_column_functor_ (const char *cn, DataFrame &d)
+        : col_name(cn), df(d)  {   }
+
+    const char  *col_name;
+    DataFrame   &df;
 
     template
     void operator() (T &vec) const;
diff --git a/include/DataFrame/Internals/DataFrame_get.tcc b/include/DataFrame/Internals/DataFrame_get.tcc
index d87250f7..1cc2e4f4 100644
--- a/include/DataFrame/Internals/DataFrame_get.tcc
+++ b/include/DataFrame/Internals/DataFrame_get.tcc
@@ -433,7 +433,8 @@ get_data_by_idx(const StlVecType &values) const  {
     const SpinGuard guard(lock_);
 
     for (const auto &citer : column_list_) [[likely]]  {
-        create_col_functor_ functor(citer.first.c_str(), df);
+        create_col_functor_ functor(citer.first.c_str(),
+                                                       df);
 
         data_[citer.second].change(functor);
     }
@@ -553,10 +554,8 @@ DataFrame::get_view_by_idx (Index2D range) const  {
         const SpinGuard guard(lock_);
 
         for (const auto &iter : column_list_) [[likely]]  {
-            view_setup_functor_  functor (iter.first.c_str(),
-                                                             b_dist,
-                                                             e_dist,
-                                                             dfcv);
+            view_setup_functor_  functor (
+                iter.first.c_str(), b_dist, e_dist, dfcv);
 
             data_[iter.second].change(functor);
         }
@@ -759,7 +758,8 @@ get_data_by_loc (const StlVecType &locations) const  {
     const SpinGuard guard(lock_);
 
     for (const auto &citer : column_list_) [[likely]]  {
-        create_col_functor_ functor(citer.first.c_str(), df);
+        create_col_functor_ functor(citer.first.c_str(),
+                                                       df);
 
         data_[citer.second].change(functor);
     }
@@ -1002,60 +1002,7 @@ get_data_by_sel (const char *name, F &sel_functor) const  {
         if (sel_functor (indices_[i], vec[i]))
             col_indices.push_back(i);
 
-    DataFrame       df;
-    IndexVecType    new_index;
-
-    new_index.reserve(col_indices.size());
-    for (const auto &citer: col_indices) [[likely]]
-        new_index.push_back(indices_[citer]);
-    df.load_index(std::move(new_index));
-
-    const SpinGuard guard(lock_);
-
-    for (const auto &citer : column_list_) [[likely]]  {
-        create_col_functor_ functor(citer.first.c_str(), df);
-
-        data_[citer.second].change(functor);
-    }
-
-    const auto  thread_level =
-        (idx_s < ThreadPool::MUL_THR_THHOLD) ? 0L : get_thread_level();
-
-    if (thread_level > 2)  {
-        auto    lbd =
-            [&col_indices = std::as_const(col_indices), idx_s, &df, this]
-            (const auto &begin, const auto &end) -> void  {
-                for (auto citer = begin; citer < end; ++citer)  {
-                    sel_load_functor_    functor (
-                        citer->first.c_str(),
-                        col_indices,
-                        idx_s,
-                        df);
-
-                    this->data_[citer->second].change(functor);
-                }
-            };
-
-        auto    futuers =
-            thr_pool_.parallel_loop(column_list_.begin(),
-                                    column_list_.end(),
-                                    std::move(lbd));
-
-        for (auto &fut : futuers)  fut.get();
-    }
-    else  {
-        for (const auto &citer : column_list_) [[likely]]  {
-            sel_load_functor_    functor (
-                citer.first.c_str(),
-                col_indices,
-                idx_s,
-                df);
-
-            data_[citer.second].change(functor);
-        }
-    }
-
-    return (df);
+    return (data_by_sel_common_(col_indices, idx_s));
 }
 
 // ----------------------------------------------------------------------------
@@ -1078,37 +1025,14 @@ get_view_by_sel (const char *name, F &sel_functor)  {
         if (sel_functor (indices_[i], vec[i])) [[unlikely]]
             col_indices.push_back(i);
 
-    using TheView = PtrView;
-
-    TheView                         dfv;
-    typename TheView::IndexVecType  new_index;
-
-    new_index.reserve(col_indices.size());
-    for (const auto &citer: col_indices) [[likely]]
-        new_index.push_back(&(indices_[citer]));
-    dfv.indices_ = std::move(new_index);
-
-    const SpinGuard guard(lock_);
-
-    for (const auto &col_citer : column_list_) [[likely]]  {
-        sel_load_view_functor_   functor (
-            col_citer.first.c_str(),
-            col_indices,
-            idx_s,
-            dfv);
-
-        data_[col_citer.second].change(functor);
-    }
-
-    return (dfv);
+    return (view_by_sel_common_(col_indices, idx_s));
 }
 
 // ----------------------------------------------------------------------------
 
 template
 template
-typename DataFrame::ConstPtrView
-DataFrame::
+typename DataFrame::ConstPtrView DataFrame::
 get_view_by_sel (const char *name, F &sel_functor) const  {
 
     static_assert(std::is_base_of, H>::value,
@@ -1124,29 +1048,7 @@ get_view_by_sel (const char *name, F &sel_functor) const  {
         if (sel_functor (indices_[i], vec[i])) [[unlikely]]
             col_indices.push_back(i);
 
-    using TheView = ConstPtrView;
-
-    TheView                         dfv;
-    typename TheView::IndexVecType  new_index;
-
-    new_index.reserve(col_indices.size());
-    for (const auto &citer: col_indices) [[likely]]
-        new_index.push_back(&(indices_[citer]));
-    dfv.indices_ = std::move(new_index);
-
-    const SpinGuard guard(lock_);
-
-    for (const auto &col_citer : column_list_) [[likely]]  {
-        sel_load_view_functor_   functor (
-            col_citer.first.c_str(),
-            col_indices,
-            idx_s,
-            dfv);
-
-        data_[col_citer.second].change(functor);
-    }
-
-    return (dfv);
+    return (view_by_sel_common_(col_indices, idx_s));
 }
 
 // ----------------------------------------------------------------------------
@@ -1175,58 +1077,7 @@ get_data_by_sel (const char *name1, const char *name2, F &sel_functor) const  {
                          i < col_s2 ? vec2[i] : get_nan()))
             col_indices.push_back(i);
 
-    DataFrame       df;
-    IndexVecType    new_index;
-
-    new_index.reserve(col_indices.size());
-    for (const auto &citer: col_indices)
-        new_index.push_back(indices_[citer]);
-    df.load_index(std::move(new_index));
-
-    for (const auto &citer : column_list_) [[likely]]  {
-        create_col_functor_ functor(citer.first.c_str(), df);
-
-        data_[citer.second].change(functor);
-    }
-
-    const auto  thread_level =
-        (idx_s < ThreadPool::MUL_THR_THHOLD) ? 0L : get_thread_level();
-
-    if (thread_level > 2)  {
-        auto    lbd =
-            [&col_indices = std::as_const(col_indices), idx_s, &df, this]
-            (const auto &begin, const auto &end) -> void  {
-                for (auto citer = begin; citer < end; ++citer)  {
-                    sel_load_functor_    functor (
-                        citer->first.c_str(),
-                        col_indices,
-                        idx_s,
-                        df);
-
-                    this->data_[citer->second].change(functor);
-                }
-            };
-
-        auto    futuers =
-            thr_pool_.parallel_loop(column_list_.begin(),
-                                    column_list_.end(),
-                                    std::move(lbd));
-
-        for (auto &fut : futuers)  fut.get();
-    }
-    else  {
-        for (const auto &citer : column_list_) [[likely]]  {
-            sel_load_functor_    functor (
-                citer.first.c_str(),
-                col_indices,
-                idx_s,
-                df);
-
-            data_[citer.second].change(functor);
-        }
-    }
-
-    return (df);
+    return (data_by_sel_common_(col_indices, idx_s));
 }
 
 // ----------------------------------------------------------------------------
@@ -1258,27 +1109,7 @@ get_view_by_sel (const char *name1, const char *name2, F &sel_functor)  {
                          i < col_s2 ? vec2[i] : get_nan()))
             col_indices.push_back(i);
 
-    using TheView = PtrView;
-
-    TheView                         dfv;
-    typename TheView::IndexVecType  new_index;
-
-    new_index.reserve(col_indices.size());
-    for (const auto &citer: col_indices)
-        new_index.push_back(&(indices_[citer]));
-    dfv.indices_ = std::move(new_index);
-
-    for (const auto &col_citer : column_list_) [[likely]]  {
-        sel_load_view_functor_   functor (
-            col_citer.first.c_str(),
-            col_indices,
-            idx_s,
-            dfv);
-
-        data_[col_citer.second].change(functor);
-    }
-
-    return (dfv);
+    return (view_by_sel_common_(col_indices, idx_s));
 }
 
 // ----------------------------------------------------------------------------
@@ -1311,27 +1142,7 @@ get_view_by_sel (const char *name1, const char *name2, F &sel_functor) const  {
                          i < col_s2 ? vec2[i] : get_nan()))
             col_indices.push_back(i);
 
-    using TheView = ConstPtrView;
-
-    TheView                         dfv;
-    typename TheView::IndexVecType  new_index;
-
-    new_index.reserve(col_indices.size());
-    for (const auto &citer: col_indices) [[likely]]
-        new_index.push_back(&(indices_[citer]));
-    dfv.indices_ = std::move(new_index);
-
-    for (const auto &col_citer : column_list_) [[likely]]  {
-        sel_load_view_functor_   functor (
-            col_citer.first.c_str(),
-            col_indices,
-            idx_s,
-            dfv);
-
-        data_[col_citer.second].change(functor);
-    }
-
-    return (dfv);
+    return (view_by_sel_common_(col_indices, idx_s));
 }
 
 // ----------------------------------------------------------------------------
@@ -1366,58 +1177,7 @@ get_data_by_sel (const char *name1,
                          i < col_s3 ? vec3[i] : get_nan()))
             col_indices.push_back(i);
 
-    DataFrame       df;
-    IndexVecType    new_index;
-
-    new_index.reserve(col_indices.size());
-    for (const auto &citer: col_indices) [[likely]]
-        new_index.push_back(indices_[citer]);
-    df.load_index(std::move(new_index));
-
-    for (const auto &citer : column_list_) [[likely]]  {
-        create_col_functor_ functor(citer.first.c_str(), df);
-
-        data_[citer.second].change(functor);
-    }
-
-    const auto  thread_level =
-        (idx_s < ThreadPool::MUL_THR_THHOLD) ? 0L : get_thread_level();
-
-    if (thread_level > 2)  {
-        auto    lbd =
-            [&col_indices = std::as_const(col_indices), idx_s, &df, this]
-            (const auto &begin, const auto &end) -> void  {
-                for (auto citer = begin; citer < end; ++citer)  {
-                    sel_load_functor_    functor (
-                        citer->first.c_str(),
-                        col_indices,
-                        idx_s,
-                        df);
-
-                    this->data_[citer->second].change(functor);
-                }
-            };
-
-        auto    futuers =
-            thr_pool_.parallel_loop(column_list_.begin(),
-                                    column_list_.end(),
-                                    std::move(lbd));
-
-        for (auto &fut : futuers)  fut.get();
-    }
-    else  {
-        for (const auto &citer : column_list_) [[likely]]  {
-            sel_load_functor_    functor (
-                citer.first.c_str(),
-                col_indices,
-                idx_s,
-                df);
-
-            data_[citer.second].change(functor);
-        }
-    }
-
-    return (df);
+    return (data_by_sel_common_(col_indices, idx_s));
 }
 
 // ----------------------------------------------------------------------------
@@ -1649,27 +1409,7 @@ get_view_by_sel (const char *name1,
                          i < col_s3 ? vec3[i] : get_nan()))
             col_indices.push_back(i);
 
-    using TheView = PtrView;
-
-    TheView                         dfv;
-    typename TheView::IndexVecType  new_index;
-
-    new_index.reserve(col_indices.size());
-    for (const auto &citer: col_indices) [[likely]]
-        new_index.push_back(&(indices_[citer]));
-    dfv.indices_ = std::move(new_index);
-
-    for (const auto &col_citer : column_list_) [[likely]]  {
-        sel_load_view_functor_   functor (
-            col_citer.first.c_str(),
-            col_indices,
-            idx_s,
-            dfv);
-
-        data_[col_citer.second].change(functor);
-    }
-
-    return (dfv);
+    return (view_by_sel_common_(col_indices, idx_s));
 }
 
 // ----------------------------------------------------------------------------
@@ -1708,27 +1448,7 @@ get_view_by_sel (const char *name1,
                          i < col_s3 ? vec3[i] : get_nan()))
             col_indices.push_back(i);
 
-    using TheView = ConstPtrView;
-
-    TheView                         dfv;
-    typename TheView::IndexVecType  new_index;
-
-    new_index.reserve(col_indices.size());
-    for (const auto &citer: col_indices) [[likely]]
-        new_index.push_back(&(indices_[citer]));
-    dfv.indices_ = std::move(new_index);
-
-    for (const auto &col_citer : column_list_) [[likely]]  {
-        sel_load_view_functor_   functor (
-            col_citer.first.c_str(),
-            col_indices,
-            idx_s,
-            dfv);
-
-        data_[col_citer.second].change(functor);
-    }
-
-    return (dfv);
+    return (view_by_sel_common_(col_indices, idx_s));
 }
 
 // ----------------------------------------------------------------------------
@@ -1770,58 +1490,7 @@ get_data_by_sel(const char *name1,
                         i < col_s4 ? vec4[i] : get_nan()))
             col_indices.push_back(i);
 
-    DataFrame       df;
-    IndexVecType    new_index;
-
-    new_index.reserve(col_indices.size());
-    for (const auto &citer: col_indices) [[likely]]
-        new_index.push_back(indices_[citer]);
-    df.load_index(std::move(new_index));
-
-    for (const auto &citer : column_list_) [[likely]]  {
-        create_col_functor_ functor(citer.first.c_str(), df);
-
-        data_[citer.second].change(functor);
-    }
-
-    const auto  thread_level =
-        (idx_s < ThreadPool::MUL_THR_THHOLD) ? 0L : get_thread_level();
-
-    if (thread_level > 2)  {
-        auto    lbd =
-            [&col_indices = std::as_const(col_indices), idx_s, &df, this]
-            (const auto &begin, const auto &end) -> void  {
-                for (auto citer = begin; citer < end; ++citer)  {
-                    sel_load_functor_    functor (
-                        citer->first.c_str(),
-                        col_indices,
-                        idx_s,
-                        df);
-
-                    this->data_[citer->second].change(functor);
-                }
-            };
-
-        auto    futuers =
-            thr_pool_.parallel_loop(column_list_.begin(),
-                                    column_list_.end(),
-                                    std::move(lbd));
-
-        for (auto &fut : futuers)  fut.get();
-    }
-    else  {
-        for (const auto &citer : column_list_) [[likely]]  {
-            sel_load_functor_    functor (
-                citer.first.c_str(),
-                col_indices,
-                idx_s,
-                df);
-
-            data_[citer.second].change(functor);
-        }
-    }
-
-    return (df);
+    return (data_by_sel_common_(col_indices, idx_s));
 }
 
 // ----------------------------------------------------------------------------
@@ -1866,27 +1535,7 @@ get_view_by_sel(const char *name1,
                         i < col_s4 ? vec4[i] : get_nan()))
             col_indices.push_back(i);
 
-    using TheView = PtrView;
-
-    TheView                         dfv;
-    typename TheView::IndexVecType  new_index;
-
-    new_index.reserve(col_indices.size());
-    for (const auto &citer: col_indices) [[likely]]
-        new_index.push_back(&(indices_[citer]));
-    dfv.indices_ = std::move(new_index);
-
-    for (const auto &col_citer : column_list_) [[likely]]  {
-        sel_load_view_functor_   functor (
-            col_citer.first.c_str(),
-            col_indices,
-            idx_s,
-            dfv);
-
-        data_[col_citer.second].change(functor);
-    }
-
-    return (dfv);
+    return (view_by_sel_common_(col_indices, idx_s));
 }
 
 // ----------------------------------------------------------------------------
@@ -1932,27 +1581,7 @@ get_view_by_sel(const char *name1,
                         i < col_s4 ? vec4[i] : get_nan()))
             col_indices.push_back(i);
 
-    using TheView = ConstPtrView;
-
-    TheView                         dfv;
-    typename TheView::IndexVecType  new_index;
-
-    new_index.reserve(col_indices.size());
-    for (const auto &citer: col_indices) [[likely]]
-        new_index.push_back(&(indices_[citer]));
-    dfv.indices_ = std::move(new_index);
-
-    for (const auto &col_citer : column_list_) [[likely]]  {
-        sel_load_view_functor_   functor (
-            col_citer.first.c_str(),
-            col_indices,
-            idx_s,
-            dfv);
-
-        data_[col_citer.second].change(functor);
-    }
-
-    return (dfv);
+    return (view_by_sel_common_(col_indices, idx_s));
 }
 
 // ----------------------------------------------------------------------------
@@ -1999,58 +1628,7 @@ get_data_by_sel(const char *name1,
                         i < col_s5 ? vec5[i] : get_nan()))
             col_indices.push_back(i);
 
-    DataFrame       df;
-    IndexVecType    new_index;
-
-    new_index.reserve(col_indices.size());
-    for (const auto &citer: col_indices) [[likely]]
-        new_index.push_back(indices_[citer]);
-    df.load_index(std::move(new_index));
-
-    for (const auto &citer : column_list_) [[likely]]  {
-        create_col_functor_ functor(citer.first.c_str(), df);
-
-        data_[citer.second].change(functor);
-    }
-
-    const auto  thread_level =
-        (idx_s < ThreadPool::MUL_THR_THHOLD) ? 0L : get_thread_level();
-
-    if (thread_level > 2)  {
-        auto    lbd =
-            [&col_indices = std::as_const(col_indices), idx_s, &df, this]
-            (const auto &begin, const auto &end) -> void  {
-                for (auto citer = begin; citer < end; ++citer)  {
-                    sel_load_functor_    functor (
-                        citer->first.c_str(),
-                        col_indices,
-                        idx_s,
-                        df);
-
-                    this->data_[citer->second].change(functor);
-                }
-            };
-
-        auto    futuers =
-            thr_pool_.parallel_loop(column_list_.begin(),
-                                    column_list_.end(),
-                                    std::move(lbd));
-
-        for (auto &fut : futuers)  fut.get();
-    }
-    else  {
-        for (const auto &citer : column_list_) [[likely]]  {
-            sel_load_functor_    functor (
-                citer.first.c_str(),
-                col_indices,
-                idx_s,
-                df);
-
-            data_[citer.second].change(functor);
-        }
-    }
-
-    return (df);
+    return (data_by_sel_common_(col_indices, idx_s));
 }
 
 // ----------------------------------------------------------------------------
@@ -2117,67 +1695,16 @@ get_data_by_sel(const char *name1,
                         i < col_s2 ? vec2[i] : get_nan(),
                         i < col_s3 ? vec3[i] : get_nan(),
                         i < col_s4 ? vec4[i] : get_nan(),
-                        i < col_s5 ? vec5[i] : get_nan(),
-                        i < col_s6 ? vec6[i] : get_nan(),
-                        i < col_s7 ? vec7[i] : get_nan(),
-                        i < col_s8 ? vec8[i] : get_nan(),
-                        i < col_s9 ? vec9[i] : get_nan(),
-                        i < col_s10 ? vec10[i] : get_nan(),
-                        i < col_s11 ? vec11[i] : get_nan()))
-            col_indices.push_back(i);
-
-    DataFrame       df;
-    IndexVecType    new_index;
-
-    new_index.reserve(col_indices.size());
-    for (const auto &citer: col_indices) [[likely]]
-        new_index.push_back(indices_[citer]);
-    df.load_index(std::move(new_index));
-
-    for (const auto &citer : column_list_) [[likely]]  {
-        create_col_functor_ functor(citer.first.c_str(), df);
-
-        data_[citer.second].change(functor);
-    }
-
-    const auto  thread_level =
-        (idx_s < ThreadPool::MUL_THR_THHOLD) ? 0L : get_thread_level();
-
-    if (thread_level > 2)  {
-        auto    lbd =
-            [&col_indices = std::as_const(col_indices), idx_s, &df, this]
-            (const auto &begin, const auto &end) -> void  {
-                for (auto citer = begin; citer < end; ++citer)  {
-                    sel_load_functor_    functor (
-                        citer->first.c_str(),
-                        col_indices,
-                        idx_s,
-                        df);
-
-                    this->data_[citer->second].change(functor);
-                }
-            };
-
-        auto    futuers =
-            thr_pool_.parallel_loop(column_list_.begin(),
-                                    column_list_.end(),
-                                    std::move(lbd));
-
-        for (auto &fut : futuers)  fut.get();
-    }
-    else  {
-        for (const auto &citer : column_list_) [[likely]]  {
-            sel_load_functor_    functor (
-                citer.first.c_str(),
-                col_indices,
-                idx_s,
-                df);
-
-            data_[citer.second].change(functor);
-        }
-    }
+                        i < col_s5 ? vec5[i] : get_nan(),
+                        i < col_s6 ? vec6[i] : get_nan(),
+                        i < col_s7 ? vec7[i] : get_nan(),
+                        i < col_s8 ? vec8[i] : get_nan(),
+                        i < col_s9 ? vec9[i] : get_nan(),
+                        i < col_s10 ? vec10[i] : get_nan(),
+                        i < col_s11 ? vec11[i] : get_nan()))
+            col_indices.push_back(i);
 
-    return (df);
+    return (data_by_sel_common_(col_indices, idx_s));
 }
 
 // ----------------------------------------------------------------------------
@@ -2257,58 +1784,7 @@ get_data_by_sel(const char *name1,
                         i < col_s12 ? vec12[i] : get_nan()))
             col_indices.push_back(i);
 
-    DataFrame       df;
-    IndexVecType    new_index;
-
-    new_index.reserve(col_indices.size());
-    for (const auto &citer: col_indices)
-        new_index.push_back(indices_[citer]);
-    df.load_index(std::move(new_index));
-
-    for (const auto &citer : column_list_) [[likely]]  {
-        create_col_functor_ functor(citer.first.c_str(), df);
-
-        data_[citer.second].change(functor);
-    }
-
-    const auto  thread_level =
-        (idx_s < ThreadPool::MUL_THR_THHOLD) ? 0L : get_thread_level();
-
-    if (thread_level > 2)  {
-        auto    lbd =
-            [&col_indices = std::as_const(col_indices), idx_s, &df, this]
-            (const auto &begin, const auto &end) -> void  {
-                for (auto citer = begin; citer < end; ++citer)  {
-                    sel_load_functor_    functor (
-                        citer->first.c_str(),
-                        col_indices,
-                        idx_s,
-                        df);
-
-                    this->data_[citer->second].change(functor);
-                }
-            };
-
-        auto    futuers =
-            thr_pool_.parallel_loop(column_list_.begin(),
-                                    column_list_.end(),
-                                    std::move(lbd));
-
-        for (auto &fut : futuers)  fut.get();
-    }
-    else  {
-        for (const auto &citer : column_list_) [[likely]]  {
-            sel_load_functor_    functor (
-                citer.first.c_str(),
-                col_indices,
-                idx_s,
-                df);
-
-            data_[citer.second].change(functor);
-        }
-    }
-
-    return (df);
+    return (data_by_sel_common_(col_indices, idx_s));
 }
 
 // ----------------------------------------------------------------------------
@@ -2392,58 +1868,7 @@ get_data_by_sel(const char *name1,
                         i < col_s13 ? vec13[i] : get_nan()))
             col_indices.push_back(i);
 
-    DataFrame       df;
-    IndexVecType    new_index;
-
-    new_index.reserve(col_indices.size());
-    for (const auto &citer: col_indices)
-        new_index.push_back(indices_[citer]);
-    df.load_index(std::move(new_index));
-
-    for (const auto &citer : column_list_) [[likely]]  {
-        create_col_functor_ functor(citer.first.c_str(), df);
-
-        data_[citer.second].change(functor);
-    }
-
-    const auto  thread_level =
-        (idx_s < ThreadPool::MUL_THR_THHOLD) ? 0L : get_thread_level();
-
-    if (thread_level > 2)  {
-        auto    lbd =
-            [&col_indices = std::as_const(col_indices), idx_s, &df, this]
-            (const auto &begin, const auto &end) -> void  {
-                for (auto citer = begin; citer < end; ++citer)  {
-                    sel_load_functor_    functor (
-                        citer->first.c_str(),
-                        col_indices,
-                        idx_s,
-                        df);
-
-                    this->data_[citer->second].change(functor);
-                }
-            };
-
-        auto    futuers =
-            thr_pool_.parallel_loop(column_list_.begin(),
-                                    column_list_.end(),
-                                    std::move(lbd));
-
-        for (auto &fut : futuers)  fut.get();
-    }
-    else  {
-        for (const auto &citer : column_list_) [[likely]]  {
-            sel_load_functor_    functor (
-                citer.first.c_str(),
-                col_indices,
-                idx_s,
-                df);
-
-            data_[citer.second].change(functor);
-        }
-    }
-
-    return (df);
+    return (data_by_sel_common_(col_indices, idx_s));
 }
 
 // ----------------------------------------------------------------------------
@@ -2492,27 +1917,7 @@ get_view_by_sel(const char *name1,
                         i < col_s5 ? vec5[i] : get_nan()))
             col_indices.push_back(i);
 
-    using TheView = PtrView;
-
-    TheView                         dfv;
-    typename TheView::IndexVecType  new_index;
-
-    new_index.reserve(col_indices.size());
-    for (const auto &citer: col_indices)
-        new_index.push_back(&(indices_[citer]));
-    dfv.indices_ = std::move(new_index);
-
-    for (const auto &col_citer : column_list_)  {
-        sel_load_view_functor_   functor (
-            col_citer.first.c_str(),
-            col_indices,
-            idx_s,
-            dfv);
-
-        data_[col_citer.second].change(functor);
-    }
-
-    return (dfv);
+    return (view_by_sel_common_(col_indices, idx_s));
 }
 
 // ----------------------------------------------------------------------------
@@ -2562,27 +1967,274 @@ get_view_by_sel(const char *name1,
                         i < col_s5 ? vec5[i] : get_nan()))
             col_indices.push_back(i);
 
-    using TheView = ConstPtrView;
+    return (view_by_sel_common_(col_indices, idx_s));
+}
 
-    TheView                         dfv;
-    typename TheView::IndexVecType  new_index;
+// ----------------------------------------------------------------------------
 
-    new_index.reserve(col_indices.size());
-    for (const auto &citer: col_indices)
-        new_index.push_back(&(indices_[citer]));
-    dfv.indices_ = std::move(new_index);
+template
+template
+DataFrame DataFrame::
+get_data_by_like(const char *name,
+                 const char *pattern,
+                 bool case_insensitive,
+                 char esc_char) const  {
 
-    for (const auto &col_citer : column_list_)  {
-        sel_load_view_functor_   functor (
-            col_citer.first.c_str(),
-            col_indices,
-            idx_s,
-            dfv);
+    const ColumnVecType  &vec = get_column(name);
+    const size_type         col_s = vec.size();
+    StlVecType   col_indices;
 
-        data_[col_citer.second].change(functor);
+    col_indices.reserve(col_s / 2);
+    for (size_type i = 0; i < col_s; ++i)  {
+        if constexpr (std::is_same_v ||
+                      std::is_same_v)  {
+            if (_like_clause_compare_(pattern,
+                                      vec[i].c_str(),
+                                      case_insensitive,
+                                      esc_char))
+                col_indices.push_back(i);
+        }
+        else  {
+            if (_like_clause_compare_(pattern,
+                                      vec[i],
+                                      case_insensitive,
+                                      esc_char))
+                col_indices.push_back(i);
+        }
     }
 
-    return (dfv);
+    return (data_by_sel_common_(col_indices, indices_.size()));
+}
+
+// ----------------------------------------------------------------------------
+
+template
+template
+typename DataFrame::PtrView DataFrame::
+get_view_by_like(const char *name,
+                 const char *pattern,
+                 bool case_insensitive,
+                 char esc_char)  {
+
+    static_assert(std::is_base_of, H>::value,
+                  "Only a StdDataFrame can call get_view_by_like()");
+
+    const ColumnVecType  &vec = get_column(name);
+    const size_type         col_s = vec.size();
+    StlVecType   col_indices;
+
+    col_indices.reserve(col_s / 2);
+    for (size_type i = 0; i < col_s; ++i) [[likely]]  {
+        if constexpr (std::is_same_v ||
+                      std::is_same_v)  {
+            if (_like_clause_compare_(pattern,
+                                      vec[i].c_str(),
+                                      case_insensitive,
+                                      esc_char))
+                col_indices.push_back(i);
+        }
+        else  {
+            if (_like_clause_compare_(pattern,
+                                      vec[i],
+                                      case_insensitive,
+                                      esc_char))
+                col_indices.push_back(i);
+        }
+    }
+
+    return (view_by_sel_common_(col_indices, indices_.size()));
+}
+
+// ----------------------------------------------------------------------------
+
+template
+template
+typename DataFrame::ConstPtrView DataFrame::
+get_view_by_like(const char *name,
+                 const char *pattern,
+                 bool case_insensitive,
+                 char esc_char) const  {
+
+    static_assert(std::is_base_of, H>::value,
+                  "Only a StdDataFrame can call get_view_by_like()");
+
+    const ColumnVecType  &vec = get_column(name);
+    const size_type         col_s = vec.size();
+    StlVecType   col_indices;
+
+    col_indices.reserve(col_s / 2);
+    for (size_type i = 0; i < col_s; ++i) [[likely]]  {
+        if constexpr (std::is_same_v ||
+                      std::is_same_v)  {
+            if (_like_clause_compare_(pattern,
+                                      vec[i].c_str(),
+                                      case_insensitive,
+                                      esc_char))
+                col_indices.push_back(i);
+        }
+        else  {
+            if (_like_clause_compare_(pattern,
+                                      vec[i],
+                                      case_insensitive,
+                                      esc_char))
+                col_indices.push_back(i);
+        }
+    }
+
+    return (view_by_sel_common_(col_indices, indices_.size()));
+}
+
+// ----------------------------------------------------------------------------
+
+template
+template
+DataFrame DataFrame::
+get_data_by_like(const char *name1,
+                 const char *name2,
+                 const char *pattern1,
+                 const char *pattern2,
+                 bool case_insensitive,
+                 char esc_char) const  {
+
+    const size_type         idx_s = indices_.size();
+    const SpinGuard         guard (lock_);
+    const ColumnVecType  &vec1 = get_column(name1, false);
+    const ColumnVecType  &vec2 = get_column(name2, false);
+    const size_type         min_col_s = std::min(vec1.size(), vec2.size());
+    StlVecType   col_indices;
+
+    col_indices.reserve(idx_s / 2);
+    for (size_type i = 0; i < min_col_s; ++i) [[likely]]  {
+        if constexpr (std::is_same_v ||
+                      std::is_same_v)  {
+            if (_like_clause_compare_(pattern1,
+                                      vec1[i].c_str(),
+                                      case_insensitive,
+                                      esc_char) &&
+                _like_clause_compare_(pattern2,
+                                      vec2[i].c_str(),
+                                      case_insensitive,
+                                      esc_char))
+                col_indices.push_back(i);
+        }
+        else  {
+            if (_like_clause_compare_(pattern1,
+                                      vec1[i],
+                                      case_insensitive,
+                                      esc_char) &&
+                _like_clause_compare_(pattern2,
+                                      vec2[i],
+                                      case_insensitive,
+                                      esc_char))
+                col_indices.push_back(i);
+        }
+    }
+
+    return (data_by_sel_common_(col_indices, idx_s));
+}
+
+// ----------------------------------------------------------------------------
+
+template
+template
+typename DataFrame::PtrView DataFrame::
+get_view_by_like(const char *name1,
+                 const char *name2,
+                 const char *pattern1,
+                 const char *pattern2,
+                 bool case_insensitive,
+                 char esc_char)  {
+
+    static_assert(std::is_base_of, H>::value,
+                  "Only a StdDataFrame can call get_view_by_like()");
+
+    const SpinGuard         guard (lock_);
+    const ColumnVecType  &vec1 = get_column(name1, false);
+    const ColumnVecType  &vec2 = get_column(name2, false);
+    const size_type         idx_s = indices_.size();
+    const size_type         min_col_s = std::min(vec1.size(), vec2.size());
+    StlVecType   col_indices;
+
+    col_indices.reserve(idx_s / 2);
+    for (size_type i = 0; i < min_col_s; ++i) [[likely]]  {
+        if constexpr (std::is_same_v ||
+                      std::is_same_v)  {
+            if (_like_clause_compare_(pattern1,
+                                      vec1[i].c_str(),
+                                      case_insensitive,
+                                      esc_char) &&
+                _like_clause_compare_(pattern2,
+                                      vec2[i].c_str(),
+                                      case_insensitive,
+                                      esc_char))
+                col_indices.push_back(i);
+        }
+        else  {
+            if (_like_clause_compare_(pattern1,
+                                      vec1[i],
+                                      case_insensitive,
+                                      esc_char) &&
+                _like_clause_compare_(pattern2,
+                                      vec2[i],
+                                      case_insensitive,
+                                      esc_char))
+                col_indices.push_back(i);
+        }
+    }
+
+    return (view_by_sel_common_(col_indices, idx_s));
+}
+
+// ----------------------------------------------------------------------------
+
+template
+template
+typename DataFrame::ConstPtrView DataFrame::
+get_view_by_like(const char *name1,
+                 const char *name2,
+                 const char *pattern1,
+                 const char *pattern2,
+                 bool case_insensitive,
+                 char esc_char) const  {
+
+    static_assert(std::is_base_of, H>::value,
+                  "Only a StdDataFrame can call get_view_by_like()");
+
+    const SpinGuard         guard (lock_);
+    const ColumnVecType  &vec1 = get_column(name1, false);
+    const ColumnVecType  &vec2 = get_column(name2, false);
+    const size_type         idx_s = indices_.size();
+    const size_type         min_col_s = std::min(vec1.size(), vec2.size());
+    StlVecType   col_indices;
+
+    col_indices.reserve(idx_s / 2);
+    for (size_type i = 0; i < min_col_s; ++i) [[likely]]  {
+        if constexpr (std::is_same_v ||
+                      std::is_same_v)  {
+            if (_like_clause_compare_(pattern1,
+                                      vec1[i].c_str(),
+                                      case_insensitive,
+                                      esc_char) &&
+                _like_clause_compare_(pattern2,
+                                      vec2[i].c_str(),
+                                      case_insensitive,
+                                      esc_char))
+                col_indices.push_back(i);
+        }
+        else  {
+            if (_like_clause_compare_(pattern1,
+                                      vec1[i],
+                                      case_insensitive,
+                                      esc_char) &&
+                _like_clause_compare_(pattern2,
+                                      vec2[i],
+                                      case_insensitive,
+                                      esc_char))
+                col_indices.push_back(i);
+        }
+    }
+
+    return (view_by_sel_common_(col_indices, idx_s));
 }
 
 // ----------------------------------------------------------------------------
@@ -3250,7 +2902,8 @@ combine(const char *col_name,
     guard.release();
 
     const size_type col_s =
-        std::min({ lhs_col.size(), df1_col.size(), df2_col.size() });
+        std::min({ lhs_col.size(), df1_col.size(),
+                              df2_col.size() });
     StlVecType   result;
 
     result.reserve(col_s);
@@ -3287,7 +2940,8 @@ combine(const char *col_name,
     result.reserve(col_s);
     for (size_type i = 0; i < col_s; ++i) [[likely]]
         result.push_back(
-            std::move(functor(lhs_col[i], df1_col[i], df2_col[i], df3_col[i])));
+            std::move(functor(lhs_col[i], df1_col[i], df2_col[i],
+                              df3_col[i])));
 
     return (result);
 }
@@ -3360,6 +3014,23 @@ DataFrame::get_str_col_stats(const char *col_name) const  {
     return (result);
 }
 
+// ----------------------------------------------------------------------------
+
+template
+template
+typename DataFrame::size_type
+DataFrame::inversion_count(const char *col_name) const  {
+
+    const auto      &col = get_column(col_name);
+    const auto      col_s = col.size();
+    StlVecType   original = col;
+    StlVecType   temp(col_s);
+    const auto      thread_level = get_thread_level();
+
+    return (_inv_merge_sort_(original, temp, 0, col_s - 1, C{ },
+                             thread_level));
+}
+
 } // namespace hmdf
 
 // ----------------------------------------------------------------------------
diff --git a/include/DataFrame/Internals/DataFrame_join.tcc b/include/DataFrame/Internals/DataFrame_join.tcc
index 42b6a20b..e8956f51 100644
--- a/include/DataFrame/Internals/DataFrame_join.tcc
+++ b/include/DataFrame/Internals/DataFrame_join.tcc
@@ -96,22 +96,18 @@ join_by_index (const RHS_T &rhs, join_policy mp) const  {
 
     switch(mp)  {
         case join_policy::inner_join:
-            return (index_inner_join_
-                        
-                    (*this, rhs, idx_vec_lhs, idx_vec_rhs));
+            return (index_inner_join_
+                        (*this, rhs, idx_vec_lhs, idx_vec_rhs));
         case join_policy::left_join:
-            return (index_left_join_
-                        
-                    (*this, rhs, idx_vec_lhs, idx_vec_rhs));
+            return (index_left_join_
+                        (*this, rhs, idx_vec_lhs, idx_vec_rhs));
         case join_policy::right_join:
-            return (index_right_join_
-                        
-                    (*this, rhs, idx_vec_lhs, idx_vec_rhs));
+            return (index_right_join_
+                        (*this, rhs, idx_vec_lhs, idx_vec_rhs));
         case join_policy::left_right_join:
         default:
-            return (index_left_right_join_
-                        
-                    (*this, rhs, idx_vec_lhs, idx_vec_rhs));
+            return (index_left_right_join_
+                        (*this, rhs, idx_vec_lhs, idx_vec_rhs));
     }
 }
 
@@ -174,161 +170,23 @@ join_by_column (const RHS_T &rhs, const char *name, join_policy mp) const  {
 
     switch(mp)  {
         case join_policy::inner_join:
-            return (column_inner_join_
-                        
-                            (*this, rhs, name, col_vec_lhs, col_vec_rhs));
+            return (column_inner_join_
+                        (*this, rhs, name, col_vec_lhs, col_vec_rhs));
         case join_policy::left_join:
-            return (column_left_join_
-                        
-                            (*this, rhs, name, col_vec_lhs, col_vec_rhs));
+            return (column_left_join_
+                        (*this, rhs, name, col_vec_lhs, col_vec_rhs));
         case join_policy::right_join:
-            return (column_right_join_
-                        
-                            (*this, rhs, name, col_vec_lhs, col_vec_rhs));
+            return (column_right_join_
+                        (*this, rhs, name, col_vec_lhs, col_vec_rhs));
         case join_policy::left_right_join:
         default:
-            return (column_left_right_join_
-                        
-                            (*this, rhs, name, col_vec_lhs, col_vec_rhs));
+            return (column_left_right_join_
+                        (*this, rhs, name, col_vec_lhs, col_vec_rhs));
     }
 }
 
 // ----------------------------------------------------------------------------
 
-template
-template
-void DataFrame::
-join_helper_common_(
-    const LHS_T &lhs,
-    const RHS_T &rhs,
-    const IndexIdxVector &joined_index_idx,
-    DataFrame> &result,
-    const char *skip_col_name)  {
-
-    using res_t = decltype(result);
-
-    std::vector>  futures;
-    const auto                      thread_level =
-        (lhs.indices_.size() < ThreadPool::MUL_THR_THHOLD)
-            ? 0L : get_thread_level();
-
-
-    if (thread_level > 2)
-        futures.reserve(lhs.column_list_.size() + rhs.column_list_.size());
-
-    const SpinGuard guard(lock_);
-
-    // NOTE: I had to do this in two separate loops. Otherwise, it would
-    //       occasionally crash in multithreaded mode under MacOS.
-    //
-    for (const auto &citer : lhs.column_list_) [[likely]]  {
-        if (skip_col_name && citer.first == skip_col_name)  continue;
-
-        if (rhs.column_tb_.find(citer.first) != rhs.column_tb_.end())  {
-            create_join_common_col_functor_   create_f(
-                citer.first.c_str(), result);
-
-            lhs.data_[citer.second].change(create_f);
-        }
-        else  {
-            create_col_functor_  create_f(
-                citer.first.c_str(), result);
-
-            lhs.data_[citer.second].change(create_f);
-        }
-    }
-
-    // Load the common and lhs columns
-    //
-    for (const auto &citer : lhs.column_list_) [[likely]]  {
-        if (skip_col_name && citer.first == skip_col_name)  continue;
-
-        // Common column between two frames
-        //
-        if (rhs.column_tb_.find(citer.first) != rhs.column_tb_.end())  {
-            auto    jcomm_lbd =
-                [&citer = std::as_const(citer),
-                 &lhs = std::as_const(lhs),
-                 &rhs = std::as_const(rhs),
-                 &joined_index_idx = std::as_const(joined_index_idx),
-                 &result] () -> void  {
-                    index_join_functor_common_   functor(
-                        citer.first.c_str(),
-                        rhs,
-                        joined_index_idx,
-                        result);
-
-                    lhs.data_[citer.second].change(functor);
-                };
-
-            if (thread_level > 2)
-                futures.emplace_back(thr_pool_.dispatch(false, jcomm_lbd));
-            else
-                jcomm_lbd();
-        }
-        else  {  // lhs only column
-            auto    jlhs_lbd =
-                [&citer = std::as_const(citer),
-                 &lhs = std::as_const(lhs),
-                 &joined_index_idx = std::as_const(joined_index_idx),
-                 &result] () -> void  {
-                    // 0 = Left
-                    index_join_functor_oneside_<0, res_t, Ts ...>   functor (
-                        citer.first.c_str(),
-                        joined_index_idx,
-                        result);
-
-                    lhs.data_[citer.second].change(functor);
-                };
-
-            if (thread_level > 2)
-                futures.emplace_back(thr_pool_.dispatch(false, jlhs_lbd));
-            else
-                jlhs_lbd();
-        }
-    }
-
-    // Load the rhs columns
-    //
-    for (const auto &citer : rhs.column_list_) [[likely]]  {
-        const auto  lhs_citer = lhs.column_tb_.find(citer.first);
-
-        if (skip_col_name && citer.first == skip_col_name)  continue;
-
-        if (lhs_citer == lhs.column_tb_.end())  {  // rhs only column
-            create_col_functor_  create_f(
-                citer.first.c_str(), result);
-
-            rhs.data_[citer.second].change(create_f);
-        }
-
-        if (lhs_citer == lhs.column_tb_.end())  {  // rhs only column
-            auto    jrhs_lbd =
-                [&citer = std::as_const(citer),
-                 &rhs = std::as_const(rhs),
-                 &joined_index_idx = std::as_const(joined_index_idx),
-                 &result] () -> void  {
-                    // 1 = Right
-                    index_join_functor_oneside_<1, res_t, Ts ...>   functor (
-                        citer.first.c_str(),
-                        joined_index_idx,
-                        result);
-
-                    rhs.data_[citer.second].change(functor);
-                };
-
-            if (thread_level > 2)
-                futures.emplace_back(thr_pool_.dispatch(false, jrhs_lbd));
-            else
-                jrhs_lbd();
-        }
-    }
-
-    for (auto &fut : futures)  fut.get();
-}
-
-// ----------------------------------------------------------------------------
-
 template
 template
 DataFrame> DataFrame::
diff --git a/include/DataFrame/Internals/DataFrame_misc.tcc b/include/DataFrame/Internals/DataFrame_misc.tcc
index 4b7b8a10..a6e3d16c 100644
--- a/include/DataFrame/Internals/DataFrame_misc.tcc
+++ b/include/DataFrame/Internals/DataFrame_misc.tcc
@@ -71,6 +71,19 @@ DataFrame::shrink_to_fit_functor_::operator() (T &vec) const  {
 
 // ----------------------------------------------------------------------------
 
+template
+template
+template
+void
+DataFrame::remove_column_functor_::operator() (T &) const  {
+
+    using value_type = typename T::value_type;
+
+    df.remove_column(col_name);
+}
+
+// ----------------------------------------------------------------------------
+
 template
 template
 template
@@ -200,7 +213,7 @@ DataFrame::print_csv_functor_::operator() (const T &vec)  {
 
     if (vec_size > 0)  {
         for (long i = sr; i < er; ++i)
-            os << vec[i] << ',';
+            _write_csv_df_index_(os, vec[i]) << ',';
     }
     os << '\n';
 
@@ -266,7 +279,7 @@ template
 void DataFrame::
 print_csv2_data_functor_::operator() (const T &vec)  {
 
-    if (vec.size() > index)  os << vec[index];
+    if (vec.size() > index)  _write_csv_df_index_(os, vec[index]);
     return;
 }
 
diff --git a/include/DataFrame/Internals/DataFrame_private_decl.h b/include/DataFrame/Internals/DataFrame_private_decl.h
index 06ccfcd5..bbc5a1d3 100644
--- a/include/DataFrame/Internals/DataFrame_private_decl.h
+++ b/include/DataFrame/Internals/DataFrame_private_decl.h
@@ -41,13 +41,17 @@ template class OPT, typename ... Ts>
 friend DF
 binary_operation(const DF &lhs, const DF &rhs);
 
-template
-size_type
-load_pair_(std::pair &col_name_data, bool do_lock = true);
+// ----------------------------------------------------------------------------
+
+// Maps row number -> number of missing column(s)
+//
+using DropRowMap = DFMap;
+using IndexIdxVector = StlVecType>;
 
 template
-size_type
-append_row_(std::pair &row_name_data);
+using JoinSortingPair = std::pair;
+
+// ----------------------------------------------------------------------------
 
 void read_json_(std::istream &file, bool columns_only);
 void read_csv_(std::istream &file, bool columns_only);
@@ -56,84 +60,10 @@ void read_csv2_(std::istream &file,
                 size_type starting_row,
                 size_type num_rows);
 
-template
-void remove_data_by_sel_common_(const StlVecType &col_indices);
-
-template
-static void
-fill_missing_value_(ColumnVecType &vec,
-                    const T &value,
-                    int limit,
-                    size_type col_num);
-
-template
-static void
-fill_missing_ffill_(ColumnVecType &vec, int limit, size_type col_num);
-
-template::value &&
-             supports_arithmetic::value>::type* = nullptr>
-static void
-fill_missing_midpoint_(ColumnVecType &vec, int limit, size_type col_num);
-
-template::value ||
-             ! supports_arithmetic::value>::type* = nullptr>
-static void
-fill_missing_midpoint_(ColumnVecType &vec, int limit, size_type col_num);
-
-template
-static void
-fill_missing_bfill_(ColumnVecType &vec, int limit);
-
-template::value &&
-             supports_arithmetic::value>::type* = nullptr>
-static void
-fill_missing_linter_(ColumnVecType &vec,
-                     const IndexVecType &index,
-                     int limit);
-
-template::value ||
-             ! supports_arithmetic::value>::type* = nullptr>
-static void
-fill_missing_linter_(ColumnVecType &, const IndexVecType &, int);
-
-// Maps row number -> number of missing column(s)
-using DropRowMap = DFMap;
-
-
-template
-static void
-drop_missing_rows_(T &vec,
-                   const DropRowMap &missing_row_map,
-                   drop_policy policy,
-                   size_type threshold,
-                   size_type col_num);
-
 template
 void
 setup_view_column_(const char *name, Index2D range);
 
-using IndexIdxVector = StlVecType>;
-template
-using JoinSortingPair = std::pair;
-
-template
-static void
-join_helper_common_(const LHS_T &lhs,
-                    const RHS_T &rhs,
-                    const IndexIdxVector &joined_index_idx,
-                    DataFrame<
-                        IDX_T,
-                        HeteroVector> &result,
-                    const char *skip_col_name = nullptr);
-
 template
 static DataFrame>
 index_join_helper_(const LHS_T &lhs,
@@ -149,16 +79,14 @@ column_join_helper_(const LHS_T &lhs,
 
 template
 static IndexIdxVector
-get_inner_index_idx_vector_(
-    const StlVecType> &col_vec_lhs,
-    const StlVecType> &col_vec_rhs);
+get_inner_index_idx_vector_(const StlVecType> &col_vec_lhs,
+                            const StlVecType> &col_vec_rhs);
 
 template
 static DataFrame>
-index_inner_join_(
-    const LHS_T &lhs, const RHS_T &rhs,
-    const StlVecType> &col_vec_lhs,
-    const StlVecType> &col_vec_rhs);
+index_inner_join_(const LHS_T &lhs, const RHS_T &rhs,
+                  const StlVecType> &col_vec_lhs,
+                  const StlVecType> &col_vec_rhs);
 
 template
 static DataFrame>
@@ -170,16 +98,14 @@ column_inner_join_(const LHS_T &lhs,
 
 template
 static IndexIdxVector
-get_left_index_idx_vector_(
-    const StlVecType> &col_vec_lhs,
-    const StlVecType> &col_vec_rhs);
+get_left_index_idx_vector_(const StlVecType> &col_vec_lhs,
+                           const StlVecType> &col_vec_rhs);
 
 template
 static DataFrame>
-index_left_join_(
-    const LHS_T &lhs, const RHS_T &rhs,
-    const StlVecType> &col_vec_lhs,
-    const StlVecType> &col_vec_rhs);
+index_left_join_(const LHS_T &lhs, const RHS_T &rhs,
+                 const StlVecType> &col_vec_lhs,
+                 const StlVecType> &col_vec_rhs);
 
 template
 static DataFrame>
@@ -191,16 +117,14 @@ column_left_join_(const LHS_T &lhs,
 
 template
 static IndexIdxVector
-get_right_index_idx_vector_(
-    const StlVecType> &col_vec_lhs,
-    const StlVecType> &col_vec_rhs);
+get_right_index_idx_vector_(const StlVecType> &col_vec_lhs,
+                            const StlVecType> &col_vec_rhs);
 
 template
 static DataFrame>
-index_right_join_(
-    const LHS_T &lhs, const RHS_T &rhs,
-    const StlVecType> &col_vec_lhs,
-    const StlVecType> &col_vec_rhs);
+index_right_join_(const LHS_T &lhs, const RHS_T &rhs,
+                  const StlVecType> &col_vec_lhs,
+                  const StlVecType> &col_vec_rhs);
 
 template
 static DataFrame>
@@ -210,13 +134,6 @@ column_right_join_(const LHS_T &lhs,
                    const StlVecType> &col_vec_lhs,
                    const StlVecType> &col_vec_rhs);
 
-template
-static DataFrame
-remove_dups_common_(const DataFrame &s_df,
-                    remove_dup_spec rds,
-                    const MAP &row_table,
-                    const IndexVecType &index);
-
 template
 static void
 concat_helper_(LHS_T &lhs, const RHS_T &rhs, bool add_new_columns);
@@ -236,12 +153,695 @@ index_left_right_join_(
 
 template
 static DataFrame>
-column_left_right_join_(
+column_left_right_join_(const LHS_T &lhs,
+                        const RHS_T &rhs,
+                        const char *col_name,
+                        const StlVecType> &col_vec_lhs,
+                        const StlVecType> &col_vec_rhs);
+
+// ----------------------------------------------------------------------------
+
+template
+size_type
+load_pair_(std::pair &col_name_data, bool do_lock = true)  {
+
+    return (load_column(
+                col_name_data.first, // column name
+                std::forward(col_name_data.second),
+                nan_policy::pad_with_nans,
+                do_lock));
+}
+
+// ----------------------------------------------------------------------------
+
+template
+size_type
+append_row_(std::pair &row_name_data)  {
+
+    return (append_column(row_name_data.first, // column name
+                             std::forward(row_name_data.second),
+                             nan_policy::dont_pad_with_nans));
+}
+
+// ----------------------------------------------------------------------------
+
+template
+static void
+drop_missing_rows_(T &vec,
+                   const DropRowMap &missing_row_map,
+                   drop_policy policy,
+                   size_type threshold,
+                   size_type col_num)  {
+
+    size_type   erase_count = 0;
+    auto        dropper =
+        [&vec, &erase_count](const auto &iter) -> void  {
+            vec.erase(vec.begin() + (iter.first - erase_count));
+            erase_count += 1;
+        };
+
+    if (policy == drop_policy::all)  {
+        for (const auto &iter : missing_row_map)
+            if (iter.second == col_num)  dropper(iter);
+    }
+    else if (policy == drop_policy::any)  {
+        for (const auto &iter : missing_row_map)
+            if (iter.second > 0)  dropper(iter);
+    }
+    else if (policy == drop_policy::threshold)  {
+        for (const auto &iter : missing_row_map)
+            if (iter.second > threshold)  dropper(iter);
+    }
+
+    return;
+}
+
+// ----------------------------------------------------------------------------
+
+template
+static void
+fill_missing_value_(ColumnVecType &vec,
+                    const T &value,
+                    int limit,
+                    size_type col_num)  {
+
+    const size_type vec_size = vec.size();
+    int             count = 0;
+
+    if (limit < 0)
+        vec.reserve(col_num);
+    for (size_type i = 0; i < col_num; ++i)  {
+        if (limit >= 0 && count >= limit)  break;
+        if (i >= vec_size)  {
+            vec.push_back(value);
+            count += 1;
+        }
+        else if (is_nan(vec[i]))  {
+            vec[i] = value;
+            count += 1;
+        }
+    }
+    return;
+}
+
+// ----------------------------------------------------------------------------
+
+template
+static void
+fill_missing_ffill_(ColumnVecType &vec, int limit, size_type col_num)  {
+
+    const size_type vec_size = vec.size();
+
+    if (vec_size == 0)  return;
+
+    int count = 0;
+    T   last_value = vec[0];
+
+    for (size_type i = 1; i < col_num; ++i)  {
+        if (limit >= 0 && count >= limit)  break;
+        if (i >= vec_size)  {
+            if (! is_nan(last_value))  {
+                vec.reserve(col_num);
+                vec.push_back(last_value);
+                count += 1;
+            }
+            else  break;
+        }
+        else  {
+            if (! is_nan(vec[i]))
+                last_value = vec[i];
+            else if (! is_nan(last_value))  {
+                vec[i] = last_value;
+                count += 1;
+            }
+        }
+    }
+    return;
+}
+
+// ----------------------------------------------------------------------------
+
+template::value &&
+             supports_arithmetic::value>::type* = nullptr>
+static void
+fill_missing_midpoint_(ColumnVecType &vec, int limit, size_type)  {
+
+    const size_type vec_size = vec.size();
+
+    if (vec_size < 3) [[unlikely]]  return;
+
+    int count = 0;
+    T   last_value = vec[0];
+
+    for (size_type i = 1; i < vec_size - 1; ++i)  {
+        if (limit >= 0 && count >= limit)  break;
+
+        if (! is_nan(vec[i]))
+            last_value = vec[i];
+        else if (! is_nan(last_value))  {
+            for (size_type j = i + 1; j < vec_size; ++j)  {
+                if (! is_nan(vec[j]))  {
+                    vec[i] = (last_value + vec[j]) / T(2);
+                    last_value = vec[i];
+                    count += 1;
+                    break;
+                }
+            }
+        }
+    }
+    return;
+}
+
+// ----------------------------------------------------------------------------
+
+template::value ||
+             ! supports_arithmetic::value>::type* = nullptr>
+static void
+fill_missing_midpoint_(ColumnVecType &, int, size_type)  {
+
+    throw NotFeasible("fill_missing_midpoint_(): ERROR: Mid-point filling is "
+                      "not feasible on non-arithmetic types");
+}
+
+// ----------------------------------------------------------------------------
+
+template
+static void
+fill_missing_bfill_(ColumnVecType &vec, int limit)  {
+
+    const long  vec_size = static_cast(vec.size());
+
+    if (vec_size == 0)  return;
+
+    int count = 0;
+    T   last_value = vec[vec_size - 1];
+
+    for (long i = vec_size - 1; i >= 0; --i)  {
+        if (limit >= 0 && count >= limit)  break;
+        if (! is_nan(vec[i]))  last_value = vec[i];
+        if (is_nan(vec[i]) && ! is_nan(last_value))  {
+            vec[i] = last_value;
+            count += 1;
+        }
+    }
+    return;
+}
+
+// ----------------------------------------------------------------------------
+
+template::value &&
+             supports_arithmetic::value>::type* = nullptr>
+static void
+fill_missing_linter_(ColumnVecType &vec,
+                     const IndexVecType &index,
+                     int limit)  {
+
+    const long  vec_size = static_cast(vec.size());
+
+    if (vec_size < 3)  return;
+
+    int             count = 0;
+    const T         *y1 = &(vec[0]);
+    const T         *y2 = &(vec[2]);
+    const IndexType *x = &(index[1]);
+    const IndexType *x1 = &(index[0]);
+    const IndexType *x2 = &(index[2]);
+
+    for (long i = 1; i < vec_size - 1; ++i)  {
+        if (limit >= 0 && count >= limit)  break;
+        if (is_nan(vec[i]))  {
+            if (is_nan(*y2))  {
+                bool    found = false;
+
+                for (long j = i + 1; j < vec_size; ++j)  {
+                    if (! is_nan(vec[j]))  {
+                        y2 = &(vec[j]);
+                        x2 = &(index[j]);
+                        found = true;
+                        break;
+                    }
+                }
+                if (! found)  break;
+            }
+            if (is_nan(*y1))  {
+                for (long j = i - 1; j >= 0; --j)  {
+                    if (! is_nan(vec[j]))  {
+                        y1 = &(vec[j]);
+                        x1 = &(index[j]);
+                        break;
+                    }
+                }
+            }
+            vec[i] =
+                *y1 +
+                (static_cast(*x - *x1) / static_cast(*x2 - *x1)) *
+                (*y2 - *y1);
+            count += 1;
+        }
+        if (i < (vec_size - 2))  {
+            y1 = &(vec[i]);
+            y2 = &(vec[i + 2]);
+            x = &(index[i + 1]);
+            x1 = &(index[i]);
+            x2 = &(index[i + 2]);
+        }
+    }
+    return;
+}
+
+// ----------------------------------------------------------------------------
+
+template::value ||
+             ! supports_arithmetic::value>::type* = nullptr>
+static void
+fill_missing_linter_(ColumnVecType &, const IndexVecType &, int)  {
+
+    throw NotFeasible("fill_missing_linter_(): ERROR: Interpolation is "
+                      "not feasible on non-arithmetic types");
+}
+
+// ----------------------------------------------------------------------------
+
+template
+void remove_data_by_sel_common_(const StlVecType &col_indices)  {
+
+    const auto  thread_level =
+        (indices_.size() < ThreadPool::MUL_THR_THHOLD)
+            ? 0L : get_thread_level();
+    SpinGuard   guard (lock_);
+
+    if (thread_level > 2)  {
+        auto    lbd =
+            [&col_indices = std::as_const(col_indices), this]
+            (const auto &begin, const auto &end) -> void  {
+                const sel_remove_functor_   functor (col_indices);
+
+                for (auto citer = begin; citer < end; ++citer)
+                    this->data_[citer->second].change(functor);
+            };
+        auto    lbd_idx =
+            [&col_indices = std::as_const(col_indices), this] () -> void  {
+                const size_type col_indices_s = col_indices.size();
+                size_type       del_count = 0;
+
+                for (size_type i = 0; i < col_indices_s; ++i) [[likely]]
+                    this->indices_.erase(this->indices_.begin() +
+                                         (col_indices[i] - del_count++));
+            };
+            auto    future_idx = thr_pool_.dispatch(false, lbd_idx);
+            auto    futures =
+                thr_pool_.parallel_loop(column_list_.begin(),
+                                        column_list_.end(),
+                                        std::move(lbd));
+
+            future_idx.get();
+            for (auto &fut : futures)  fut.get();
+    }
+    else  {
+        const sel_remove_functor_   functor (col_indices);
+
+        for (const auto &citer : column_list_)
+            data_[citer.second].change(functor);
+        guard.release();
+
+        const size_type col_indices_s = col_indices.size();
+        size_type       del_count = 0;
+
+        for (size_type i = 0; i < col_indices_s; ++i) [[likely]]
+            indices_.erase(indices_.begin() + (col_indices[i] - del_count++));
+    }
+
+    return;
+}
+
+// ----------------------------------------------------------------------------
+
+template
+static void
+join_helper_common_(
     const LHS_T &lhs,
     const RHS_T &rhs,
-    const char *col_name,
-    const StlVecType> &col_vec_lhs,
-    const StlVecType> &col_vec_rhs);
+    const IndexIdxVector &joined_index_idx,
+    DataFrame> &result,
+    const char *skip_col_name = nullptr)  {
+
+    using res_t = decltype(result);
+
+    std::vector>  futures;
+    const auto                      thread_level =
+        (lhs.indices_.size() < ThreadPool::MUL_THR_THHOLD)
+            ? 0L : get_thread_level();
+
+
+    if (thread_level > 2)
+        futures.reserve(lhs.column_list_.size() + rhs.column_list_.size());
+
+    const SpinGuard guard(lock_);
+
+    // NOTE: I had to do this in two separate loops. Otherwise, it would
+    //       occasionally crash in multithreaded mode under MacOS.
+    //
+    for (const auto &citer : lhs.column_list_) [[likely]]  {
+        if (skip_col_name && citer.first == skip_col_name)  continue;
+
+        if (rhs.column_tb_.find(citer.first) != rhs.column_tb_.end())  {
+            create_join_common_col_functor_   create_f(
+                citer.first.c_str(), result);
+
+            lhs.data_[citer.second].change(create_f);
+        }
+        else  {
+            create_col_functor_  create_f(
+                citer.first.c_str(), result);
+
+            lhs.data_[citer.second].change(create_f);
+        }
+    }
+
+    // Load the common and lhs columns
+    //
+    for (const auto &citer : lhs.column_list_) [[likely]]  {
+        if (skip_col_name && citer.first == skip_col_name)  continue;
+
+        // Common column between two frames
+        //
+        if (rhs.column_tb_.find(citer.first) != rhs.column_tb_.end())  {
+            auto    jcomm_lbd =
+                [&citer = std::as_const(citer),
+                 &lhs = std::as_const(lhs),
+                 &rhs = std::as_const(rhs),
+                 &joined_index_idx = std::as_const(joined_index_idx),
+                 &result] () -> void  {
+                    index_join_functor_common_   functor(
+                        citer.first.c_str(),
+                        rhs,
+                        joined_index_idx,
+                        result);
+
+                    lhs.data_[citer.second].change(functor);
+                };
+
+            if (thread_level > 2)
+                futures.emplace_back(thr_pool_.dispatch(false, jcomm_lbd));
+            else
+                jcomm_lbd();
+        }
+        else  {  // lhs only column
+            auto    jlhs_lbd =
+                [&citer = std::as_const(citer),
+                 &lhs = std::as_const(lhs),
+                 &joined_index_idx = std::as_const(joined_index_idx),
+                 &result] () -> void  {
+                    // 0 = Left
+                    index_join_functor_oneside_<0, res_t, Ts ...>   functor (
+                        citer.first.c_str(),
+                        joined_index_idx,
+                        result);
+
+                    lhs.data_[citer.second].change(functor);
+                };
+
+            if (thread_level > 2)
+                futures.emplace_back(thr_pool_.dispatch(false, jlhs_lbd));
+            else
+                jlhs_lbd();
+        }
+    }
+
+    // Load the rhs columns
+    //
+    for (const auto &citer : rhs.column_list_) [[likely]]  {
+        const auto  lhs_citer = lhs.column_tb_.find(citer.first);
+
+        if (skip_col_name && citer.first == skip_col_name)  continue;
+
+        if (lhs_citer == lhs.column_tb_.end())  {  // rhs only column
+            create_col_functor_  create_f(
+                citer.first.c_str(), result);
+
+            rhs.data_[citer.second].change(create_f);
+        }
+
+        if (lhs_citer == lhs.column_tb_.end())  {  // rhs only column
+            auto    jrhs_lbd =
+                [&citer = std::as_const(citer),
+                 &rhs = std::as_const(rhs),
+                 &joined_index_idx = std::as_const(joined_index_idx),
+                 &result] () -> void  {
+                    // 1 = Right
+                    index_join_functor_oneside_<1, res_t, Ts ...>   functor (
+                        citer.first.c_str(),
+                        joined_index_idx,
+                        result);
+
+                    rhs.data_[citer.second].change(functor);
+                };
+
+            if (thread_level > 2)
+                futures.emplace_back(thr_pool_.dispatch(false, jrhs_lbd));
+            else
+                jrhs_lbd();
+        }
+    }
+
+    for (auto &fut : futures)  fut.get();
+}
+
+// ----------------------------------------------------------------------------
+
+template
+static DataFrame
+remove_dups_common_(const DataFrame &s_df,
+                    remove_dup_spec rds,
+                    const MAP &row_table,
+                    const IndexVecType &index)  {
+
+    using count_vec = StlVecType;
+
+    count_vec   rows_to_del;
+
+    rows_to_del.reserve(8);
+    if (rds == remove_dup_spec::keep_first)  {
+        for (const auto &citer : row_table)  {
+            if (citer.second.size() > 1)  {
+                for (size_type i = 1; i < citer.second.size(); ++i)
+                    rows_to_del.push_back(citer.second[i]);
+            }
+        }
+    }
+    else if (rds == remove_dup_spec::keep_last)  {
+        for (const auto &citer : row_table)  {
+            if (citer.second.size() > 1)  {
+                for (size_type i = 0; i < citer.second.size() - 1; ++i)
+                    rows_to_del.push_back(citer.second[i]);
+            }
+        }
+    }
+    else  {  // remove_dup_spec::keep_none
+        for (const auto &citer : row_table)  {
+            if (citer.second.size() > 1)  {
+                for (size_type i = 0; i < citer.second.size(); ++i)
+                    rows_to_del.push_back(citer.second[i]);
+            }
+        }
+    }
+
+    DataFrame new_df;
+    IndexVecType    new_index (index.size() - rows_to_del.size());
+    const SpinGuard guard(lock_);
+
+    // Load the index
+    //
+    _remove_copy_if_(index.begin(), index.end(), new_index.begin(),
+                     [&rows_to_del = std::as_const(rows_to_del)]
+                     (std::size_t n) -> bool  {
+                         return (std::find(rows_to_del.begin(),
+                                           rows_to_del.end(),
+                                           n) != rows_to_del.end());
+                     });
+    new_df.load_index(std::move(new_index));
+
+    // Create the columns, so loading can proceed in parallel
+    //
+    for (const auto &citer : s_df.column_list_)  {
+        create_col_functor_ functor(
+            citer.first.c_str(), new_df);
+
+        s_df.data_[citer.second].change(functor);
+    }
+
+    const auto  thread_level =
+        (new_df.get_index().size() < ThreadPool::MUL_THR_THHOLD)
+            ? 0L : get_thread_level();
+
+    if (thread_level > 2)  {
+        auto    lbd =
+            [&rows_to_del = std::as_const(rows_to_del),
+             &s_df = std::as_const(s_df),
+             &new_df]
+            (const auto &begin, const auto &end) -> void  {
+                for (auto citer = begin; citer < end; ++citer)  {
+                    copy_remove_functor_ functor (citer->first.c_str(),
+                                                          rows_to_del,
+                                                          new_df);
+
+                    s_df.data_[citer->second].change(functor);
+                }
+            };
+        auto    futures =
+            thr_pool_.parallel_loop(s_df.column_list_.begin(),
+                                    s_df.column_list_.end(),
+                                    std::move(lbd));
+
+        for (auto &fut : futures)  fut.get();
+    }
+    else  {
+        for (const auto &citer : s_df.column_list_)  {
+            copy_remove_functor_    functor (citer.first.c_str(),
+                                                     rows_to_del,
+                                                     new_df);
+
+            s_df.data_[citer.second].change(functor);
+        }
+    }
+    return (new_df);
+}
+
+// ----------------------------------------------------------------------------
+
+template
+DataFrame
+data_by_sel_common_(const StlVecType &col_indices,
+                    size_type idx_s) const  {
+
+    DataFrame       ret_df;
+    IndexVecType    new_index;
+
+    new_index.reserve(col_indices.size());
+    for (const auto &citer: col_indices) [[likely]]
+        new_index.push_back(indices_[citer]);
+    ret_df.load_index(std::move(new_index));
+
+    const SpinGuard guard(lock_);
+
+    for (const auto &citer : column_list_) [[likely]]  {
+        create_col_functor_  functor(citer.first.c_str(),
+                                                        ret_df);
+
+        data_[citer.second].change(functor);
+    }
+
+    const auto  thread_level =
+        (idx_s < ThreadPool::MUL_THR_THHOLD) ? 0L : get_thread_level();
+
+    if (thread_level > 2)  {
+        auto    lbd =
+            [&col_indices = std::as_const(col_indices), idx_s, &ret_df, this]
+            (const auto &begin, const auto &end) -> void  {
+                for (auto citer = begin; citer < end; ++citer)  {
+                    sel_load_functor_    functor (
+                        citer->first.c_str(),
+                        col_indices,
+                        idx_s,
+                        ret_df);
+
+                    this->data_[citer->second].change(functor);
+                }
+            };
+
+        auto    futuers =
+            thr_pool_.parallel_loop(column_list_.begin(),
+                                    column_list_.end(),
+                                    std::move(lbd));
+
+        for (auto &fut : futuers)  fut.get();
+    }
+    else  {
+        for (const auto &citer : column_list_) [[likely]]  {
+            sel_load_functor_    functor (
+                citer.first.c_str(),
+                col_indices,
+                idx_s,
+                ret_df);
+
+            data_[citer.second].change(functor);
+        }
+    }
+
+    return (ret_df);
+}
+
+// ----------------------------------------------------------------------------
+
+template
+PtrView
+view_by_sel_common_(const StlVecType &col_indices,
+                    size_type idx_s)  {
+
+    using TheView = PtrView;
+
+    TheView                         ret_dfv;
+    typename TheView::IndexVecType  new_index;
+
+    new_index.reserve(col_indices.size());
+    for (const auto &citer: col_indices) [[likely]]
+        new_index.push_back(&(indices_[citer]));
+    ret_dfv.indices_ = std::move(new_index);
+
+    const SpinGuard guard(lock_);
+
+    for (const auto &col_citer : column_list_) [[likely]]  {
+        sel_load_view_functor_   functor (
+            col_citer.first.c_str(),
+            col_indices,
+            idx_s,
+            ret_dfv);
+
+        data_[col_citer.second].change(functor);
+    }
+
+    return (ret_dfv);
+}
+
+// ----------------------------------------------------------------------------
+
+template
+ConstPtrView
+view_by_sel_common_(const StlVecType &col_indices,
+                    size_type idx_s) const  {
+
+    using TheView = ConstPtrView;
+
+    TheView                         ret_dfv;
+    typename TheView::IndexVecType  new_index;
+
+    new_index.reserve(col_indices.size());
+    for (const auto &citer: col_indices) [[likely]]
+        new_index.push_back(&(indices_[citer]));
+    ret_dfv.indices_ = std::move(new_index);
+
+    const SpinGuard guard(lock_);
+
+    for (const auto &col_citer : column_list_) [[likely]]  {
+        sel_load_view_functor_   functor (
+            col_citer.first.c_str(),
+            col_indices,
+            idx_s,
+            ret_dfv);
+
+        data_[col_citer.second].change(functor);
+    }
+
+    return (ret_dfv);
+}
 
 // ----------------------------------------------------------------------------
 
@@ -297,6 +897,27 @@ col_vector_push_back_func_(V &vec,
 
 // ----------------------------------------------------------------------------
 
+template
+inline static void
+col_vector_push_back_cont_func_(V &vec,
+                                std::istream &file,
+                                T (*converter)(const char *))  {
+
+    std::string value;
+    char        c = 0;
+
+    value.reserve(2048);
+    while (file.get(c)) [[likely]] {
+        value.clear();
+        if (c == '\n')  break;
+        file.unget();
+        _get_token_from_file_(file, ',', value, '\0');
+        vec.push_back(converter(value.c_str()));
+    }
+}
+
+// ----------------------------------------------------------------------------
+
 template
 struct  ColVectorPushBack_  {
 
diff --git a/include/DataFrame/Internals/DataFrame_read.tcc b/include/DataFrame/Internals/DataFrame_read.tcc
index 8871d7f3..f869ffeb 100644
--- a/include/DataFrame/Internals/DataFrame_read.tcc
+++ b/include/DataFrame/Internals/DataFrame_read.tcc
@@ -212,9 +212,46 @@ void DataFrame::read_json_(std::istream &stream, bool columns_only)  {
 
                 vec.reserve(col_size);
                 col_vector_push_back_func_(vec,
-                                            stream,
-                                            &::strtoul,
-                                            io_format::json);
+                                           stream,
+                                           &::strtoul,
+                                           io_format::json);
+            }
+            else if (col_type == "char")  {
+                StlVecType    &vec =
+                    create_column(col_name.c_str(), false);
+
+                vec.reserve(col_size);
+                col_vector_push_back_func_>(
+                    vec,
+                    stream,
+                    [](const char *tok, char **, int) -> char  {
+                        if (tok[0] == '\0')
+                            return ('\0');
+                        else if (tok[1] == '\0')
+                            return (static_cast(int(tok[0])));
+                        else
+                            return (static_cast(atoi(tok)));
+                    },
+                    io_format::json);
+            }
+            else if (col_type == "uchar")  {
+                StlVecType   &vec =
+                    create_column(col_name.c_str(), false);
+
+                vec.reserve(col_size);
+                col_vector_push_back_func_>(
+                    vec,
+                    stream,
+                    [](const char *tok, char **, int) -> unsigned char  {
+                        if (tok[0] == '\0')
+                            return ('\0');
+                        else if (tok[1] == '\0')
+                            return (static_cast(int(tok[0])));
+                        else
+                            return (static_cast(atoi(tok)));
+                    },
+                    io_format::json);
             }
             else if (col_type == "long")  {
                 StlVecType   &vec =
@@ -267,7 +304,9 @@ void DataFrame::read_json_(std::istream &stream, bool columns_only)  {
                 StlVecType   &vec =
                     create_column(col_name.c_str(), false);
                 auto                    converter =
-                    [](const char *, char **)-> DateTime { return DateTime(); };
+                    [](const char *, char **) -> DateTime {
+                        return DateTime();
+                    };
                 const ColVectorPushBack_>  slug;
 
                 vec.reserve(col_size);
@@ -396,6 +435,41 @@ void DataFrame::read_csv_(std::istream &stream, bool columns_only)  {
                 vec.reserve(::atoi(value.c_str()));
                 col_vector_push_back_func_(vec, stream, &::strtoul);
             }
+            else if (type_str == "char")  {
+                StlVecType    &vec =
+                    create_column(col_name.c_str(), false);
+
+                vec.reserve(::atoi(value.c_str()));
+                col_vector_push_back_func_>(
+                    vec,
+                    stream,
+                    [](const char *tok, char **, int) -> char  {
+                        if (tok[0] == '\0')
+                            return ('\0');
+                        else if (tok[1] == '\0')
+                            return (static_cast(int(tok[0])));
+                        else
+                            return (static_cast(atoi(tok)));
+                    });
+            }
+            else if (type_str == "uchar")  {
+                StlVecType   &vec =
+                    create_column(col_name.c_str(), false);
+
+                vec.reserve(::atoi(value.c_str()));
+                col_vector_push_back_func_>(
+                    vec,
+                    stream,
+                    [](const char *tok, char **, int) -> unsigned char  {
+                        if (tok[0] == '\0')
+                            return ('\0');
+                        else if (tok[1] == '\0')
+                            return (static_cast(int(tok[0])));
+                        else
+                            return (static_cast(atoi(tok)));
+                    });
+            }
             else if (type_str == "long")  {
                 StlVecType   &vec =
                     create_column(col_name.c_str(), false);
@@ -439,7 +513,9 @@ void DataFrame::read_csv_(std::istream &stream, bool columns_only)  {
                 StlVecType   &vec =
                     create_column(col_name.c_str(), false);
                 auto                    converter =
-                    [](const char *, char **)-> DateTime { return DateTime(); };
+                    [](const char *, char **) -> DateTime {
+                        return DateTime();
+                    };
                 const ColVectorPushBack_>  slug;
 
                 vec.reserve(::atoi(value.c_str()));
@@ -452,6 +528,79 @@ void DataFrame::read_csv_(std::istream &stream, bool columns_only)  {
                 vec.reserve(::atoi(value.c_str()));
                 col_vector_push_back_func_(vec, stream, &::strtol);
             }
+
+            // Containers
+            //
+            else if (type_str == "dbl_vec")  {
+                using vec_t = std::vector;
+
+                StlVecType   &vec =
+                    create_column(col_name.c_str(), false);
+
+                vec.reserve(::atoi(value.c_str()));
+                col_vector_push_back_cont_func_(
+                    vec,
+                    stream,
+                    &_get_dbl_vec_from_value_);
+            }
+            else if (type_str == "str_vec")  {
+                using vec_t = std::vector;
+
+                StlVecType   &vec =
+                    create_column(col_name.c_str(), false);
+
+                vec.reserve(::atoi(value.c_str()));
+                col_vector_push_back_cont_func_(
+                    vec,
+                    stream,
+                    &_get_str_vec_from_value_);
+            }
+            else if (type_str == "dbl_set")  {
+                using set_t = std::set;
+
+                StlVecType   &vec =
+                    create_column(col_name.c_str(), false);
+
+                vec.reserve(::atoi(value.c_str()));
+                col_vector_push_back_cont_func_(vec,
+                                                stream,
+                                                &_get_dbl_set_from_value_);
+            }
+            else if (type_str == "str_set")  {
+                using set_t = std::set;
+
+                StlVecType   &vec =
+                    create_column(col_name.c_str(), false);
+
+                vec.reserve(::atoi(value.c_str()));
+                col_vector_push_back_cont_func_(vec,
+                                                stream,
+                                                &_get_str_set_from_value_);
+            }
+            else if (type_str == "str_dbl_map")  {
+                using map_t = std::map;
+
+                StlVecType   &vec =
+                    create_column(col_name.c_str(), false);
+
+                vec.reserve(::atoi(value.c_str()));
+                col_vector_push_back_cont_func_(
+                    vec,
+                    stream,
+                    &_get_str_dbl_map_from_value_);
+            }
+            else if (type_str == "str_dbl_unomap")  {
+                using map_t = std::unordered_map;
+
+                StlVecType   &vec =
+                    create_column(col_name.c_str(), false);
+
+                vec.reserve(::atoi(value.c_str()));
+                col_vector_push_back_cont_func_(
+                    vec,
+                    stream,
+                    &_get_str_dbl_map_from_value_);
+            }
             else [[unlikely]]
                 throw DataFrameError("DataFrame::read_csv_(): ERROR: Unknown "
                                      "column type");
@@ -568,6 +717,16 @@ read_csv2_(std::istream &stream,
                                       type_str.c_str(),
                                       col_name.c_str(),
                                       nrows);
+            else if (type_str == "char")
+                spec_vec.emplace_back(StlVecType(),
+                                      type_str.c_str(),
+                                      col_name.c_str(),
+                                      nrows);
+            else if (type_str == "uchar")
+                spec_vec.emplace_back(StlVecType(),
+                                      type_str.c_str(),
+                                      col_name.c_str(),
+                                      nrows);
             else if (type_str == "long")
                 spec_vec.emplace_back(StlVecType(),
                                       type_str.c_str(),
@@ -605,13 +764,15 @@ read_csv2_(std::istream &stream,
                                       type_str.c_str(),
                                       col_name.c_str(),
                                       nrows);
+            // Containers
+            //
             else if (type_str == "dbl_vec")
-                spec_vec.emplace_back(StlVecType>{ },
+                spec_vec.emplace_back(StlVecType>{ },
                                       type_str.c_str(),
                                       col_name.c_str(),
                                       nrows);
             else if (type_str == "str_vec")
-                spec_vec.emplace_back(StlVecType>{ },
+                spec_vec.emplace_back(StlVecType>{ },
                                       type_str.c_str(),
                                       col_name.c_str(),
                                       nrows);
@@ -676,6 +837,25 @@ read_csv2_(std::istream &stream,
                         (col_spec.col_vec).push_back(
                             (unsigned int) strtoul(value.c_str(), nullptr, 0));
             }
+            else if (col_spec.type_spec == "char")  {
+                if (value.size() > 1)
+                    std::any_cast &>
+                        (col_spec.col_vec).push_back(
+                            static_cast(atoi(value.c_str())));
+                else if (! value.empty())
+                    std::any_cast &>
+                        (col_spec.col_vec).push_back(value[0]);
+            }
+            else if (col_spec.type_spec == "uchar")  {
+                if (value.size() > 1)
+                    std::any_cast &>
+                        (col_spec.col_vec).push_back(
+                            static_cast(atoi(value.c_str())));
+                else if (! value.empty())
+                    std::any_cast &>
+                        (col_spec.col_vec).push_back(
+                            static_cast(value[0]));
+            }
             else if (col_spec.type_spec == "long")  {
                 if (! value.empty())
                     std::any_cast &>
@@ -750,26 +930,27 @@ read_csv2_(std::istream &stream,
                     vec.push_back(v);
                 }
             }
+
+            // Containers
+            //
             else if (col_spec.type_spec == "dbl_vec")  {
                 if (! value.empty())  {
-                    StlVecType>  &vec =
-                        std::any_cast> &>
+                    StlVecType>  &vec =
+                        std::any_cast> &>
                             (col_spec.col_vec);
 
                     vec.push_back(
-                        std::move(_get_dbl_vec_from_value_>(
-                                      value.c_str())));
+                        std::move(_get_dbl_vec_from_value_(value.c_str())));
                 }
             }
             else if (col_spec.type_spec == "str_vec")  {
                 if (! value.empty())  {
-                    StlVecType> &vec =
-                        std::any_cast> &>
+                    StlVecType> &vec =
+                        std::any_cast> &>
                             (col_spec.col_vec);
 
                     vec.push_back(
-                        std::move(_get_str_vec_from_value_>(
-                                      value.c_str())));
+                        std::move(_get_str_vec_from_value_(value.c_str())));
                 }
             }
             else if (col_spec.type_spec == "dbl_set")  {
@@ -801,8 +982,9 @@ read_csv2_(std::istream &stream,
                     StlVecType   &vec =
                         std::any_cast &>(col_spec.col_vec);
 
-                    vec.push_back(std::move(_get_str_dbl_map_from_value_(
-                                      value.c_str())));
+                    vec.push_back(
+                        std::move(_get_str_dbl_map_from_value_(
+                        value.c_str())));
                 }
             }
             else if (col_spec.type_spec == "str_dbl_unomap")  {
@@ -812,8 +994,9 @@ read_csv2_(std::istream &stream,
                     StlVecType   &vec =
                         std::any_cast &> (col_spec.col_vec);
 
-                    vec.push_back(std::move(_get_str_dbl_map_from_value_(
-                                      value.c_str())));
+                    vec.push_back(
+                        std::move(_get_str_dbl_map_from_value_(
+                        value.c_str())));
                 }
             }
             col_index += 1;
@@ -867,6 +1050,18 @@ read_csv2_(std::istream &stream,
                     std::move(std::any_cast &>
                         (col_spec.col_vec)),
                     nan_policy::dont_pad_with_nans);
+            else if (col_spec.type_spec == "char")
+                load_column(
+                    col_spec.col_name.c_str(),
+                    std::move(std::any_cast &>
+                        (col_spec.col_vec)),
+                    nan_policy::dont_pad_with_nans);
+            else if (col_spec.type_spec == "uchar")
+                load_column(
+                    col_spec.col_name.c_str(),
+                    std::move(std::any_cast &>
+                        (col_spec.col_vec)),
+                    nan_policy::dont_pad_with_nans);
             else if (col_spec.type_spec == "long")
                 load_column(
                     col_spec.col_name.c_str(),
@@ -909,17 +1104,20 @@ read_csv2_(std::istream &stream,
                     std::move(std::any_cast &>
                         (col_spec.col_vec)),
                     nan_policy::dont_pad_with_nans);
+
+            // Containers
+            //
             else if (col_spec.type_spec == "dbl_vec")
-                load_column>(
+                load_column>(
                     col_spec.col_name.c_str(),
-                    std::move(std::any_cast> &>
+                    std::move(std::any_cast> &>
                         (col_spec.col_vec)),
                     nan_policy::dont_pad_with_nans);
             else if (col_spec.type_spec == "str_vec")
-                load_column>(
+                load_column>(
                     col_spec.col_name.c_str(),
                     std::move(
-                        std::any_cast> &>
+                        std::any_cast> &>
                         (col_spec.col_vec)),
                     nan_policy::dont_pad_with_nans);
             else if (col_spec.type_spec == "dbl_set")  {
diff --git a/include/DataFrame/Internals/DataFrame_set.tcc b/include/DataFrame/Internals/DataFrame_set.tcc
index 7b95aa50..1b224292 100644
--- a/include/DataFrame/Internals/DataFrame_set.tcc
+++ b/include/DataFrame/Internals/DataFrame_set.tcc
@@ -70,30 +70,22 @@ DataFrame::create_column (const char *name, bool do_lock)  {
 // ----------------------------------------------------------------------------
 
 template
+template
 void DataFrame::remove_column (const char *name)  {
 
     static_assert(std::is_base_of, DataVec>::value,
                   "Only a StdDataFrame can call remove_column()");
 
-    if (! ::strcmp(name, DF_INDEX_COL_NAME))
-        throw DataFrameError ("DataFrame::remove_column(): ERROR: "
-                              "Data column name cannot be 'INDEX'");
-
-    const auto  iter = column_tb_.find (name);
-
-    if (iter == column_tb_.end())  {
-        char    buffer [512];
+    ColumnVecType    &vec = get_column(name);
 
-        snprintf (buffer, sizeof(buffer) - 1,
-                  "DataFrame::remove_column(): ERROR: Cannot find column '%s'",
-                  name);
-        throw ColNotFound (buffer);
-    }
+    // Free the memory space
+    //
+    vec = std::move(ColumnVecType{ });
 
     // I do not erase the column from the data_ vector, because it will mess up
     // indices in the hash table column_tb_
     /* data_.erase (data_.begin() + iter->second); */
-    column_tb_.erase (iter);
+    column_tb_.erase (name);
     for (size_type i = 0; i < column_list_.size(); ++i)  {
         if (column_list_[i].first == name)  {
             column_list_.erase(column_list_.begin() + i);
@@ -107,9 +99,42 @@ void DataFrame::remove_column (const char *name)  {
 // ----------------------------------------------------------------------------
 
 template
+template
 void DataFrame::remove_column(size_type index)  {
 
-    return (remove_column(column_list_[index].first.c_str()));
+    return (remove_column(column_list_[index].first.c_str()));
+}
+
+// ----------------------------------------------------------------------------
+
+template
+void DataFrame::clear()  {
+
+    {
+        const SpinGuard guard(lock_);
+
+        data_.clear();
+    }
+    indices_.clear();
+    column_tb_.clear();
+    column_list_.clear();
+    return;
+}
+
+// ----------------------------------------------------------------------------
+
+template
+void DataFrame::swap(DataFrame &other)  {
+
+    {
+        const SpinGuard guard(lock_);
+
+        data_.swap(other.data_);
+    }
+    indices_.swap(other.indices_);
+    column_tb_.swap(other.column_tb_);
+    column_list_.swap(other.column_list_);
+    return;
 }
 
 // ----------------------------------------------------------------------------
@@ -147,7 +172,13 @@ void DataFrame::rename_column (const char *from, const char *to)  {
 
     column_tb_.emplace (to, from_iter->second);
     column_list_.emplace_back (to, from_iter->second);
-    remove_column(from);
+    column_tb_.erase (from);
+    for (size_type i = 0; i < column_list_.size(); ++i)  {
+        if (column_list_[i].first == from)  {
+            column_list_.erase(column_list_.begin() + i);
+            break;
+        }
+    }
     return;
 }
 
@@ -172,7 +203,7 @@ retype_column (const char *name,
     new_vec.reserve(old_vec.size());
     for (const auto &citer : old_vec)
         new_vec.push_back(std::move(convert_func(citer)));
-    remove_column(name);
+    remove_column(name);
     load_column(name, std::move(new_vec));
     return;
 }
@@ -431,11 +462,12 @@ load_result_as_column(V &visitor,
 
     size_type   ret_cnt = data_s;
 
-    if (padding == nan_policy::pad_with_nans && data_s < idx_s)
+    if (padding == nan_policy::pad_with_nans && data_s < idx_s)  {
         for (size_type i = 0; i < idx_s - data_s; ++i)  {
             new_col.push_back (std::move(get_nan()));
             ret_cnt += 1;
         }
+    }
 
     const auto              iter = column_tb_.find (new_col_name);
     StlVecType    *vec_ptr = nullptr;
@@ -595,13 +627,15 @@ from_indicators(const StlVecType &ind_col_names,
 
     guard.release();
     new_col.reserve(col_s);
-    for (size_type i = 0; i < col_s; ++i) [[likely]]
-        for (size_type j = 0; j < ind_col_s; ++j) [[likely]]
+    for (size_type i = 0; i < col_s; ++i) [[likely]]  {
+        for (size_type j = 0; j < ind_col_s; ++j) [[likely]]  {
             if (ind_cols[j]->at(i))  {
                 new_col.push_back(
                     _string_to_(ind_col_names[j] + pre_offset));
                 break;
             }
+        }
+    }
 
     return (col_s);
 }
@@ -770,21 +804,6 @@ load_column (const char *name,
 
 // ----------------------------------------------------------------------------
 
-template
-template
-typename DataFrame::size_type
-DataFrame::
-load_pair_(std::pair &col_name_data, bool do_lock)  {
-
-    return (load_column(
-                col_name_data.first, // column name
-                std::forward(col_name_data.second),
-                nan_policy::pad_with_nans,
-                do_lock));
-}
-
-// ----------------------------------------------------------------------------
-
 template
 template
 typename DataFrame::size_type
@@ -872,18 +891,6 @@ append_column (const char *name, const T &val, nan_policy padding)  {
 
 // ----------------------------------------------------------------------------
 
-template
-template
-typename DataFrame::size_type
-DataFrame::append_row_(std::pair &row_name_data)  {
-
-    return (append_column(row_name_data.first, // column name
-                             std::forward(row_name_data.second),
-                             nan_policy::dont_pad_with_nans));
-}
-
-// ----------------------------------------------------------------------------
-
 template
 template
 typename DataFrame::size_type
@@ -1026,62 +1033,6 @@ void DataFrame::remove_data_by_loc (Index2D range)  {
 
 // ----------------------------------------------------------------------------
 
-template
-template
-void DataFrame::
-remove_data_by_sel_common_(const StlVecType &col_indices)  {
-
-    const auto  thread_level =
-        (indices_.size() < ThreadPool::MUL_THR_THHOLD)
-            ? 0L : get_thread_level();
-    SpinGuard   guard (lock_);
-
-    if (thread_level > 2)  {
-        auto    lbd =
-            [&col_indices = std::as_const(col_indices), this]
-            (const auto &begin, const auto &end) -> void  {
-                const sel_remove_functor_   functor (col_indices);
-
-                for (auto citer = begin; citer < end; ++citer)
-                    this->data_[citer->second].change(functor);
-            };
-        auto    lbd_idx =
-            [&col_indices = std::as_const(col_indices), this] () -> void  {
-                const size_type col_indices_s = col_indices.size();
-                size_type       del_count = 0;
-
-                for (size_type i = 0; i < col_indices_s; ++i) [[likely]]
-                    this->indices_.erase(this->indices_.begin() +
-                                         (col_indices[i] - del_count++));
-            };
-            auto    future_idx = thr_pool_.dispatch(false, lbd_idx);
-            auto    futures =
-                thr_pool_.parallel_loop(column_list_.begin(),
-                                        column_list_.end(),
-                                        std::move(lbd));
-
-            future_idx.get();
-            for (auto &fut : futures)  fut.get();
-    }
-    else  {
-        const sel_remove_functor_   functor (col_indices);
-
-        for (const auto &citer : column_list_)
-            data_[citer.second].change(functor);
-        guard.release();
-
-        const size_type col_indices_s = col_indices.size();
-        size_type       del_count = 0;
-
-        for (size_type i = 0; i < col_indices_s; ++i) [[likely]]
-            indices_.erase(indices_.begin() + (col_indices[i] - del_count++));
-    }
-
-    return;
-}
-
-// ----------------------------------------------------------------------------
-
 template
 template
 void DataFrame::remove_data_by_sel (const char *name, F &sel_functor)  {
@@ -1170,107 +1121,6 @@ remove_data_by_sel (const char *name1,
 
 // ----------------------------------------------------------------------------
 
-template
-template
-DataFrame DataFrame::
-remove_dups_common_(const DataFrame &s_df,
-                    remove_dup_spec rds,
-                    const MAP &row_table,
-                    const IndexVecType &index)  {
-
-    using count_vec = StlVecType;
-
-    count_vec   rows_to_del;
-
-    rows_to_del.reserve(8);
-    if (rds == remove_dup_spec::keep_first)  {
-        for (const auto &citer : row_table)  {
-            if (citer.second.size() > 1)  {
-                for (size_type i = 1; i < citer.second.size(); ++i)
-                    rows_to_del.push_back(citer.second[i]);
-            }
-        }
-    }
-    else if (rds == remove_dup_spec::keep_last)  {
-        for (const auto &citer : row_table)  {
-            if (citer.second.size() > 1)  {
-                for (size_type i = 0; i < citer.second.size() - 1; ++i)
-                    rows_to_del.push_back(citer.second[i]);
-            }
-        }
-    }
-    else  {  // remove_dup_spec::keep_none
-        for (const auto &citer : row_table)  {
-            if (citer.second.size() > 1)  {
-                for (size_type i = 0; i < citer.second.size(); ++i)
-                    rows_to_del.push_back(citer.second[i]);
-            }
-        }
-    }
-
-    DataFrame new_df;
-    IndexVecType    new_index (index.size() - rows_to_del.size());
-    const SpinGuard guard(lock_);
-
-    // Load the index
-    //
-    _remove_copy_if_(index.begin(), index.end(), new_index.begin(),
-                     [&rows_to_del = std::as_const(rows_to_del)]
-                     (std::size_t n) -> bool  {
-                         return (std::find(rows_to_del.begin(),
-                                           rows_to_del.end(),
-                                           n) != rows_to_del.end());
-                     });
-    new_df.load_index(std::move(new_index));
-
-    // Create the columns, so loading can proceed in parallel
-    //
-    for (const auto &citer : s_df.column_list_)  {
-        create_col_functor_ functor(
-            citer.first.c_str(), new_df);
-
-        s_df.data_[citer.second].change(functor);
-    }
-
-    const auto  thread_level =
-        (new_df.get_index().size() < ThreadPool::MUL_THR_THHOLD)
-            ? 0L : get_thread_level();
-
-    if (thread_level > 2)  {
-        auto    lbd =
-            [&rows_to_del = std::as_const(rows_to_del),
-             &s_df = std::as_const(s_df),
-             &new_df]
-            (const auto &begin, const auto &end) -> void  {
-                for (auto citer = begin; citer < end; ++citer)  {
-                    copy_remove_functor_ functor (citer->first.c_str(),
-                                                          rows_to_del,
-                                                          new_df);
-
-                    s_df.data_[citer->second].change(functor);
-                }
-            };
-        auto    futures =
-            thr_pool_.parallel_loop(s_df.column_list_.begin(),
-                                    s_df.column_list_.end(),
-                                    std::move(lbd));
-
-        for (auto &fut : futures)  fut.get();
-    }
-    else  {
-        for (const auto &citer : s_df.column_list_)  {
-            copy_remove_functor_    functor (citer.first.c_str(),
-                                                     rows_to_del,
-                                                     new_df);
-
-            s_df.data_[citer.second].change(functor);
-        }
-    }
-    return (new_df);
-}
-
-// ----------------------------------------------------------------------------
-
 template
 template
 DataFrame DataFrame::
@@ -1324,12 +1174,12 @@ remove_duplicates (const char *name1,
 
     guard.release();
 
-    const auto              &index = get_index();
-    const size_type         col_s =
+    const auto      &index = get_index();
+    const size_type col_s =
         std::min({ vec1.size(), vec2.size(), index.size() });
-    map_t                   row_table;
-    count_vec               dummy_vec;
-    const IndexType         dummy_idx { };
+    map_t           row_table;
+    count_vec       dummy_vec;
+    const IndexType dummy_idx { };
 
     for (size_type i = 0; i < col_s; ++i) [[likely]]  {
         const auto  insert_res =
@@ -1374,9 +1224,9 @@ remove_duplicates (const char *name1,
     const size_type col_s =
         std::min(
             { vec1.size(), vec2.size(), vec3.size(), index.size() });
-    map_t                   row_table;
-    count_vec               dummy_vec;
-    const IndexType         dummy_idx { };
+    map_t           row_table;
+    count_vec       dummy_vec;
+    const IndexType dummy_idx { };
 
     for (size_type i = 0; i < col_s; ++i) [[likely]]  {
         const auto  insert_res =
@@ -1426,9 +1276,9 @@ remove_duplicates (const char *name1,
         std::min(
             { vec1.size(), vec2.size(), vec3.size(), vec4.size(),
               index.size() });
-    map_t                   row_table;
-    count_vec               dummy_vec;
-    const IndexType         dummy_idx { };
+    map_t           row_table;
+    count_vec       dummy_vec;
+    const IndexType dummy_idx { };
 
     for (size_type i = 0; i < col_s; ++i) [[likely]]  {
         const auto  insert_res =
@@ -1480,9 +1330,9 @@ remove_duplicates (const char *name1,
         std::min(
             { vec1.size(), vec2.size(), vec3.size(), vec4.size(), vec5.size(),
               index.size() });
-    map_t                   row_table;
-    count_vec               dummy_vec;
-    const IndexType         dummy_idx { };
+    map_t           row_table;
+    count_vec       dummy_vec;
+    const IndexType dummy_idx { };
 
     for (size_type i = 0; i < col_s; ++i) [[likely]]  {
         const auto  insert_res =
@@ -1539,9 +1389,9 @@ remove_duplicates (const char *name1,
             { vec1.size(), vec2.size(), vec3.size(), vec4.size(),
               vec5.size(), vec6.size(),
               index.size() });
-    map_t                   row_table;
-    count_vec               dummy_vec;
-    const IndexType         dummy_idx { };
+    map_t           row_table;
+    count_vec       dummy_vec;
+    const IndexType dummy_idx { };
 
     for (size_type i = 0; i < col_s; ++i) [[likely]]  {
         const auto  insert_res =
@@ -1587,8 +1437,8 @@ consolidate(const char *old_col_name1,
                        false);
     guard.release();
     if (delete_old_cols)  {
-        remove_column(old_col_name1);
-        remove_column(old_col_name2);
+        remove_column(old_col_name1);
+        remove_column(old_col_name2);
     }
     return;
 }
@@ -1626,9 +1476,9 @@ consolidate(const char *old_col_name1,
                        false);
     guard.release();
     if (delete_old_cols)  {
-        remove_column(old_col_name1);
-        remove_column(old_col_name2);
-        remove_column(old_col_name3);
+        remove_column(old_col_name1);
+        remove_column(old_col_name2);
+        remove_column(old_col_name3);
     }
     return;
 }
@@ -1670,10 +1520,10 @@ consolidate(const char *old_col_name1,
                        false);
     guard.release();
     if (delete_old_cols)  {
-        remove_column(old_col_name1);
-        remove_column(old_col_name2);
-        remove_column(old_col_name3);
-        remove_column(old_col_name4);
+        remove_column(old_col_name1);
+        remove_column(old_col_name2);
+        remove_column(old_col_name3);
+        remove_column(old_col_name4);
     }
     return;
 }
@@ -1720,11 +1570,11 @@ consolidate(const char *old_col_name1,
                        false);
     guard.release();
     if (delete_old_cols)  {
-        remove_column(old_col_name1);
-        remove_column(old_col_name2);
-        remove_column(old_col_name3);
-        remove_column(old_col_name4);
-        remove_column(old_col_name5);
+        remove_column(old_col_name1);
+        remove_column(old_col_name2);
+        remove_column(old_col_name3);
+        remove_column(old_col_name4);
+        remove_column(old_col_name5);
     }
     return;
 }
diff --git a/include/DataFrame/Internals/DataFrame_shift.tcc b/include/DataFrame/Internals/DataFrame_shift.tcc
index 7cf308b9..bcaa2520 100644
--- a/include/DataFrame/Internals/DataFrame_shift.tcc
+++ b/include/DataFrame/Internals/DataFrame_shift.tcc
@@ -65,18 +65,26 @@ void DataFrame::self_shift(size_type periods, shift_policy sp)  {
 
                 for (auto &fut : futuers)  fut.get();
             }
-            else {
+            else  {
                 for (size_type idx = 0; idx < num_cols; ++idx) [[likely]]
                     data_[idx].change(functor);
             }
         }
-        else if (sp == shift_policy::left)  {
-            while (periods-- > 0)
-                remove_column(column_list_.front().first.c_str());
-        }
-        else if (sp == shift_policy::right)  {
-            while (periods-- > 0)
-                remove_column(column_list_.back().first.c_str());
+        else  {
+            while (periods-- > 0)  {
+                const char                      *col_name =
+                    (sp == shift_policy::left)
+                        ? column_list_.front().first.c_str()
+                        : column_list_.back().first.c_str();
+                remove_column_functor_  functor (col_name, *this);
+
+                for (const auto &citer : column_list_)  {
+                    if (citer.first == col_name)  {
+                        data_[citer.second].change(functor);
+                        break;
+                    }
+                }
+            }
         }
     }
 }
diff --git a/include/DataFrame/Internals/DataFrame_standalone.tcc b/include/DataFrame/Internals/DataFrame_standalone.tcc
index ed164f4f..a95290a1 100644
--- a/include/DataFrame/Internals/DataFrame_standalone.tcc
+++ b/include/DataFrame/Internals/DataFrame_standalone.tcc
@@ -32,6 +32,7 @@ SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
 #include 
 #include 
 
+#include 
 #include 
 #include 
 #include 
@@ -70,7 +71,7 @@ static const
 std::unordered_map<_TypeInfoRef_,
                    const char *const,
                    _TypeinfoHasher_,
-                   _TypeinfoEqualTo_>   _typeinfo_name_ {
+                   _TypeinfoEqualTo_>   _typeinfo_name_  {
     { typeid(float), "float" },
     { typeid(double), "double" },
     { typeid(long double), "longdouble" },
@@ -82,9 +83,16 @@ std::unordered_map<_TypeInfoRef_,
     { typeid(unsigned long int), "ulong" },
     { typeid(long long int), "longlong" },
     { typeid(unsigned long long int), "ulonglong" },
+    { typeid(char), "char" },
+    { typeid(unsigned char), "uchar" },
     { typeid(std::string), "string" },
+    { typeid(const char *), "string" },
+    { typeid(char *), "string" },
     { typeid(bool), "bool" },
     { typeid(DateTime), "DateTime" },
+
+    // Containers
+    //
     { typeid(std::vector), "dbl_vec" },
     { typeid(std::vector), "str_vec" },
     { typeid(std::set), "dbl_set" },
@@ -277,7 +285,8 @@ _load_groupby_data_2_(
 
     std::size_t         marker = 0;
     auto                &dst_idx = dest.get_index();
-    const std::size_t   vec_size = std::min(input_col1.size(), input_col2.size());
+    const std::size_t   vec_size =
+        std::min(input_col1.size(), input_col2.size());
     const auto          &src_idx = source.get_index();
 
     if (dst_idx.empty())  {
@@ -577,6 +586,28 @@ inline static S &_write_json_df_index_(S &o, const std::string &value)  {
 
 // ----------------------------------------------------------------------------
 
+template
+inline static S &_write_json_df_index_(S &o, char value)  {
+
+    if (std::isprint(value))
+        return (o << value);
+    else
+        return (o << static_cast(value));
+}
+
+// ----------------------------------------------------------------------------
+
+template
+inline static S &_write_json_df_index_(S &o, unsigned char value)  {
+
+    if (std::isprint(value))
+        return (o << value);
+    else
+        return (o << static_cast(value));
+}
+
+// ----------------------------------------------------------------------------
+
 inline static void
 _get_token_from_file_ (std::istream &file,
                        char delim,
@@ -601,11 +632,10 @@ _get_token_from_file_ (std::istream &file,
 
 // ----------------------------------------------------------------------------
 
-template
-inline static typename DF::template StlVecType
+inline static std::vector
 _get_dbl_vec_from_value_(const char *value)  {
 
-    using vec_t = typename DF::template StlVecType;
+    using vec_t = std::vector;
 
     std::size_t vcnt = 0;
     char        buffer[128];
@@ -634,11 +664,10 @@ _get_dbl_vec_from_value_(const char *value)  {
 
 // ----------------------------------------------------------------------------
 
-template
-inline static typename DF::template StlVecType
+inline static std::vector
 _get_str_vec_from_value_(const char *value)  {
 
-    using vec_t = typename DF::template StlVecType;
+    using vec_t = std::vector;
 
     std::size_t vcnt { 0 };
     char        buffer[2048];
@@ -825,6 +854,28 @@ inline static S &_write_csv_df_index_(S &o, const DateTime &value)  {
 
 // ----------------------------------------------------------------------------
 
+template
+inline static S &_write_csv_df_index_(S &o, char value)  {
+
+    if (std::isprint(value))
+        return (o << value);
+    else
+        return (o << static_cast(value));
+}
+
+// ----------------------------------------------------------------------------
+
+template
+inline static S &_write_csv_df_index_(S &o, unsigned char value)  {
+
+    if (std::isprint(value))
+        return (o << value);
+    else
+        return (o << static_cast(value));
+}
+
+// ----------------------------------------------------------------------------
+
 //
 // Specializing std::hash for tuples
 //
@@ -935,6 +986,348 @@ inline static T _string_to_(const char *value)  {
     return (ret);
 }
 
+// ----------------------------------------------------------------------------
+
+template
+static std::size_t
+_inv_merge_(Con &original,
+            Con &temp,
+            std::size_t left,
+            std::size_t mid,
+            std::size_t right,
+            Comp &&comp)  {
+
+    std::size_t i { left };
+    std::size_t j { mid };
+    std::size_t k { left };
+    std::size_t inv_count { 0 };
+
+    while ((i <= mid - 1) && (j <= right)) {
+        if (comp (original[i], original[j])) {
+            temp[k++] = original[i++];
+        }
+        else {
+            temp[k++] = original[j++];
+            inv_count += mid - i;
+        }
+    }
+
+    // Copy the remaining elements of left sub-original (if there are any)
+    // to temp
+    //
+    while (i <= mid - 1)
+        temp[k++] = original[i++];
+
+    // Copy the remaining elements of right sub-original (if there are any)
+    // to temp
+    //
+    while (j <= right)
+        temp[k++] = original[j++];
+
+    // Copy back the merged elements to original original
+    //
+    for (i = left; i <= right; i++)
+        original[i] = temp[i];
+
+    return (inv_count);
+}
+
+// ----------------------------------------------------------------------------
+
+template
+static std::size_t
+_inv_merge_sort_(Con &original,
+               Con &temp,
+               std::size_t left,
+               std::size_t right,
+               Comp comp,
+               long thread_level)  {
+
+    using fut_type = std::future;
+
+    std::size_t mid { 0 };
+    std::size_t inv_count { 0 };
+
+    if (right > left) {
+        const auto  thr_lvl =
+            ((right - left) < (ThreadPool::MUL_THR_THHOLD / 2))
+                ? 0L : thread_level;
+
+        // Divide the original into two parts and call _inv_merge_sort_()
+        // for each of the parts
+        //
+        mid = (right + left) / 2;
+
+        // Inversion count will be sum of inversions in left-part, right-part
+        // and number of inversions in merging
+        //
+        if (thr_lvl > 2)  {
+            fut_type    left_fut =
+                ThreadGranularity::thr_pool_.dispatch(
+                    false,
+                        _inv_merge_sort_,
+                    std::ref(original),
+                    std::ref(temp),
+                    left,
+                    mid,
+                    comp,
+                    thread_level);
+            fut_type    right_fut =
+                ThreadGranularity::thr_pool_.dispatch(
+                    false,
+                    _inv_merge_sort_,
+                    std::ref(original),
+                    std::ref(temp),
+                    mid + 1,
+                    right,
+                    comp,
+                    thread_level);
+
+            ThreadGranularity::thr_pool_.run_task();
+            ThreadGranularity::thr_pool_.run_task();
+            inv_count += left_fut.get() + right_fut.get();
+        }
+        else  {
+            inv_count +=
+                _inv_merge_sort_(original, temp, left, mid, comp, thread_level);
+            inv_count +=
+                _inv_merge_sort_(original, temp, mid + 1, right, comp,
+                                 thread_level);
+        }
+
+        // Merge the two parts
+        //
+        inv_count += _inv_merge_(original, temp, left, mid + 1, right, comp);
+    }
+
+    return (inv_count);
+}
+
+// ----------------------------------------------------------------------------
+
+struct _LikeClauseUtil_  {
+
+    // This lookup table is used to help decode the first byte of
+    // a multi-byte UTF8 character.
+    //
+    inline static const unsigned char   CHAR_TRANS1[] = {
+        0x00, 0x01, 0x02, 0x03, 0x04, 0x05, 0x06, 0x07, 0x08, 0x09, 0x0a, 0x0b,
+        0x0c, 0x0d, 0x0e, 0x0f, 0x10, 0x11, 0x12, 0x13, 0x14, 0x15, 0x16, 0x17,
+        0x18, 0x19, 0x1a, 0x1b, 0x1c, 0x1d, 0x1e, 0x1f, 0x00, 0x01, 0x02, 0x03,
+        0x04, 0x05, 0x06, 0x07, 0x08, 0x09, 0x0a, 0x0b, 0x0c, 0x0d, 0x0e, 0x0f,
+        0x00, 0x01, 0x02, 0x03, 0x04, 0x05, 0x06, 0x07, 0x00, 0x01, 0x02, 0x03,
+        0x00, 0x01, 0x00, 0x00,
+    };
+
+    inline static const unsigned char   UPPER_TO_LOWER[] = {
+        0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19,
+        20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37,
+        38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55,
+        56, 57, 58, 59, 60, 61, 62, 63, 64, 97, 98, 99, 100, 101, 102, 103, 104,
+        105, 106, 107, 108, 109, 110, 111, 112, 113, 114, 115, 116, 117, 118,
+        119, 120, 121, 122, 91, 92, 93, 94, 95, 96, 97, 98, 99, 100, 101, 102,
+        103, 104, 105, 106, 107, 108, 109, 110, 111, 112, 113, 114, 115, 116,
+        117, 118, 119, 120, 121, 122, 123, 124, 125, 126, 127, 128, 129, 130,
+        131, 132, 133, 134, 135, 136, 137, 138, 139, 140, 141, 142, 143, 144,
+        145, 146, 147, 148, 149, 150, 151, 152, 153, 154, 155, 156, 157, 158,
+        159, 160, 161, 162, 163, 164, 165, 166, 167, 168, 169, 170, 171, 172,
+        173, 174, 175, 176, 177, 178, 179, 180, 181, 182, 183, 184, 185, 186,
+        187, 188, 189, 190, 191, 192, 193, 194, 195, 196, 197, 198, 199, 200,
+        201, 202, 203, 204, 205, 206, 207, 208, 209, 210, 211, 212, 213, 214,
+        215, 216, 217, 218, 219, 220, 221, 222, 223, 224, 225, 226, 227, 228,
+        229, 230, 231, 232, 233, 234, 235, 236, 237, 238, 239, 240, 241, 242,
+        243, 244, 245, 246, 247, 248, 249, 250, 251, 252, 253, 254, 255,
+    };
+
+    static unsigned int char_read(const unsigned char **str_ptr_ptr)  {
+
+        unsigned int    c = *((*str_ptr_ptr)++);
+
+        // For this routine, we assume the char string is always
+        // zero-terminated.
+        //
+        if (c >= 0xc0)  {
+            c = CHAR_TRANS1[c - 0xc0];
+            while ((*(*str_ptr_ptr) & 0xc0) == 0x80)
+                c = (c << 6) + (0x3f & *((*str_ptr_ptr)++));
+            if (c < 0x80 ||
+                (c & 0xFFFFF800) == 0xD800 || (c & 0xFFFFFFFE) == 0xFFFE)
+                c = 0xFFFD;
+        }
+
+        return (c);
+    }
+
+    static inline void upper_to_lower(unsigned int &val)  {
+
+        if (! (val & ~0x7f))
+            val = UPPER_TO_LOWER[val];
+    }
+};
+
+// ----------------------------------------------------------------------------
+
+// This compares two null-terminated strings for equality where the first
+// string can potentially be a "glob" expression (forget about regular
+// expressions).
+// It returns true if they are matched-equal and false if they are not
+// matched-equal.
+//
+// Globbing rules:
+//
+//      '*'       Matches any sequence of zero or more characters.
+//
+//      '?'       Matches exactly one character.
+//
+//     [...]      Matches one character from the enclosed list of
+//                characters.
+//
+//     [^...]     Matches one character not in the enclosed list.
+//
+// With the [...] and [^...] matching, a ']' character can be included
+// in the list by making it the first character after '[' or '^'.  A
+// range of characters can be specified using '-'.  Example:
+// "[a-z]" matches any single lower-case letter. To match a '-', make
+// it the last character in the list.
+//
+// Hints: to match '*' or '?', put them in "[]". Like this:
+//        abc[*]xyz matches "abc*xyz" only
+//
+// NOTE: This could be, in some cases, n-squared. But it is pretty fast with
+//       moderately sized strings. I have not tested this with huge/massive
+//       strings. 
+//
+static inline bool
+_like_clause_compare_(const char *pattern,
+                      const char *input_str,
+                      bool case_insensitive = false,
+                      unsigned int esc_char = '\\')  {
+
+    const unsigned char *upattern =
+        reinterpret_cast(pattern);
+    const unsigned char *uinput_str =
+        reinterpret_cast(input_str);
+    unsigned int        c, c2;
+    const unsigned char match_one { '?' };
+    const unsigned char match_all { '*' };
+    const unsigned char match_set { '[' };
+    // True if the previous character was escape
+    //
+    bool                prev_escape { false };
+
+    while ((c = _LikeClauseUtil_::char_read(&upattern)) != 0)  {
+        if (c == match_all && ! prev_escape)  {
+            while ((c = _LikeClauseUtil_::char_read(&upattern)) == match_all ||
+                   c == match_one)  {
+                if (c == match_one &&
+                    _LikeClauseUtil_::char_read(&uinput_str) == 0)
+                    return (false);
+            }
+            if (c == 0)  {
+                return (true);
+            }
+            else if (c == esc_char)  {
+                c = _LikeClauseUtil_::char_read(&upattern);
+                if (c == 0)
+                    return (false);
+            }
+            else if (c == match_set)  {
+                while (*uinput_str &&
+                       _like_clause_compare_(
+                           reinterpret_cast(&upattern[-1]),
+                           reinterpret_cast(uinput_str),
+                           case_insensitive,
+                           esc_char) == 0)  {
+                    if ((*(uinput_str++)) >= 0xc0)
+                        while ((*uinput_str & 0xc0) == 0x80)
+                            uinput_str++;
+                }
+                return (*uinput_str != 0);
+            }
+            while ((c2 = _LikeClauseUtil_::char_read(&uinput_str)) != 0)  {
+                if (case_insensitive)  {
+                    _LikeClauseUtil_::upper_to_lower(c2);
+                    _LikeClauseUtil_::upper_to_lower(c);
+                    while (c2 != 0 && c2 != c)  {
+                        c2 = _LikeClauseUtil_::char_read(&uinput_str);
+                        _LikeClauseUtil_::upper_to_lower(c2);
+                    }
+                }
+                else  {
+                    while (c2 != 0 && c2 != c)
+                        c2 = _LikeClauseUtil_::char_read(&uinput_str);
+                }
+                if (c2 == 0)
+                    return (false);
+                if (_like_clause_compare_(
+                        reinterpret_cast(upattern),
+                        reinterpret_cast(uinput_str),
+                        case_insensitive,
+                        esc_char))
+                    return (true);
+            }
+            return (false);
+        }
+        else if (c == match_one && ! prev_escape)  {
+            if (_LikeClauseUtil_::char_read(&uinput_str) == 0)
+                return (false);
+        }
+        else if (c == match_set)  {
+            unsigned int    prior_c { 0 };
+            int             seen { 0 };
+            int             invert { 0 };
+
+            c = _LikeClauseUtil_::char_read(&uinput_str);
+            if (c == 0)
+                return (false);
+            c2 = _LikeClauseUtil_::char_read(&upattern);
+            if (c2 == '^')  {
+                invert = 1;
+                c2 = _LikeClauseUtil_::char_read(&upattern);
+            }
+            if (c2 == ']')  {
+                if (c == ']')
+                    seen = 1;
+                c2 = _LikeClauseUtil_::char_read(&upattern);
+            }
+            while (c2 && c2 != ']')  {
+                if (c2 == '-' &&
+                    upattern[0] != ']' &&
+                    upattern[0] != 0 &&
+                    prior_c > 0)  {
+                    c2 = _LikeClauseUtil_::char_read(&upattern);
+                    if (c >= prior_c && c <= c2)
+                        seen = 1;
+                    prior_c = 0;
+                }
+                else  {
+                    if (c == c2)
+                        seen = 1;
+                    prior_c = c2;
+                }
+                c2 = _LikeClauseUtil_::char_read(&upattern);
+            }
+            if (c2 == 0 || (seen ^ invert) == 0)
+                return (false);
+        }
+        else if (esc_char == c && ! prev_escape)  {
+            prev_escape = true;
+        }
+        else  {
+            c2 = _LikeClauseUtil_::char_read(&uinput_str);
+            if (case_insensitive)  {
+                _LikeClauseUtil_::upper_to_lower(c);
+                _LikeClauseUtil_::upper_to_lower(c2);
+            }
+            if (c != c2)
+                return (false);
+            prev_escape = false;
+        }
+    }
+
+    return (*uinput_str == 0);
+}
+
 } // namespace hmdf
 
 // ----------------------------------------------------------------------------
diff --git a/test/dataframe_tester.cc b/test/dataframe_tester.cc
index 70f8561b..79393e06 100644
--- a/test/dataframe_tester.cc
+++ b/test/dataframe_tester.cc
@@ -516,6 +516,12 @@ static void test_read()  {
                   int,
                   unsigned long,
                   double,
+                  std::map,
+                  std::unordered_map,
+                  std::vector,
+                  std::set,
+                  std::set,
+                  std::vector,
                   std::string,
                   bool>(std::cout);
 
@@ -532,6 +538,14 @@ static void test_read()  {
                       unsigned long,
                       double,
                       std::string,
+                      char,
+                      unsigned char,
+                      std::map,
+                      std::unordered_map,
+                      std::vector,
+                      std::set,
+                      std::set,
+                      std::vector,
                       bool>(std::cout);
 
     StdDataFrame  df_read_dt;
@@ -642,7 +656,7 @@ static void test_remove_column()  {
     StlVecType            i1 = { 22, 23, 24, 25 };
     StlVecType    s1 =
         { "11", "22", "33", "xx", "yy", "gg", "string" };
-    MyDataFrame                 df;
+    MyDataFrame                df;
 
     df.load_data(std::move(idx),
                  std::make_pair("col_1", d1),
@@ -652,10 +666,10 @@ static void test_remove_column()  {
                  std::make_pair("col_str", s1));
 
     df.write(std::cout);
-    df.remove_column("col_2");
+    df.remove_column("col_2");
     std::cout << "After removing column `col_2`" << std::endl;
     df.write(std::cout);
-    df.remove_column("col_str");
+    df.remove_column("col_str");
     std::cout << "After removing column `col_str`" << std::endl;
     df.write(std::cout);
 
@@ -1262,6 +1276,8 @@ static void test_dataframe_friend_plus_operator()  {
                 int,
                 double,
                 std::string,
+                char,
+                unsigned char,
                 bool>(df1, df2);
 
     std::cout << "Original DF1:" << std::endl;
@@ -1270,6 +1286,8 @@ static void test_dataframe_friend_plus_operator()  {
               unsigned long,
               double,
               std::string,
+              char,
+              unsigned char,
               bool>(std::cout);
     std::cout << "Original DF2:" << std::endl;
     df2.write(std::cout);
     std::cout << "Result DF:" << std::endl;
     result.write(std::cout);
 }
 
@@ -1311,6 +1333,8 @@ static void test_dataframe_friend_minus_operator()  {
                  unsigned long,
                  int,
                  double,
+                 char,
+                 unsigned char,
                  bool>(df1, df2);
 
     std::cout << "Original DF1:" << std::endl;
@@ -1319,6 +1343,8 @@ static void test_dataframe_friend_minus_operator()  {
               unsigned long,
               double,
               std::string,
+              char,
+              unsigned char,
               bool>(std::cout);
     std::cout << "Original DF2:" << std::endl;
     df2.write(std::cout);
 }
 
@@ -3333,6 +3361,13 @@ static void test_reading_writing_json()  {
         assert(df.get_column("col_3").size() == 12);
         assert(df.get_column("col_3")[3] == 18.0);
         assert(df.get_column("col_3")[11] == 555.543);
+
+        assert(df.get_column("col_char")[11] == 78);
+        assert(df.get_column("col_char")[2] == 'F');
+        assert(df.get_column("col_char")[8] == 'h');
+        assert(df.get_column("col_uchar")[11] == 255);
+        assert(df.get_column("col_uchar")[7] == 88);
+        assert(df.get_column("col_uchar")[9] == '&');
     }
     catch (const DataFrameError &ex)  {
         std::cout << ex.what() << std::endl;
diff --git a/test/dataframe_tester_2.cc b/test/dataframe_tester_2.cc
index 73b30126..e72021c5 100644
--- a/test/dataframe_tester_2.cc
+++ b/test/dataframe_tester_2.cc
@@ -3105,6 +3105,8 @@ static void test_no_index_reads()  {
                  unsigned long,
                  double,
                  bool,
+                 char,
+                 unsigned char,
                  std::string>(std::cout, io_format::csv2);
 
         std::cout << '\n' << std::endl;
@@ -3116,6 +3118,14 @@ static void test_no_index_reads()  {
                   unsigned long,
                   double,
                   bool,
+                  char,
+                  unsigned char,
+                  std::map,
+                  std::unordered_map,
+                  std::vector,
+                  std::set,
+                  std::set,
+                  std::vector,
                   std::string>(std::cout, io_format::csv2);
 
         std::cout << '\n' << std::endl;
@@ -3123,10 +3133,9 @@ static void test_no_index_reads()  {
         df3.read("data/sample_data_2.json", io_format::json, true);
         df3.read("data/sample_data_no_index.json", io_format::json, true);
         df3.write(std::cout, io_format::csv2);
     }
     catch (const DataFrameError &ex)  {
diff --git a/test/dataframe_tester_3.cc b/test/dataframe_tester_3.cc
index 272252d6..08e73477 100644
--- a/test/dataframe_tester_3.cc
+++ b/test/dataframe_tester_3.cc
@@ -293,14 +293,14 @@ static void test_to_from_string()  {
           123457, 123458, 123459, 123460, 123461, 123462, 123466 };
     StlVecType d1 = { 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14 };
     StlVecType d2 = { 8, 9, 10, 11, 12, 13, 14, 20, 22, 23,
-                               30, 31, 32, 1.89 };
+                              30, 31, 32, 1.89 };
     StlVecType d3 = { 15, 16, 17, 18, 19, 20, 21,
-                               0.34, 1.56, 0.34, 2.3, 0.1, 0.89, 0.45 };
+                              0.34, 1.56, 0.34, 2.3, 0.1, 0.89, 0.45 };
     StlVecType    i1 = { 22, 23, 24, 25, 99, 100, 101, 3, 2 };
     StlVecType    strvec =
         { "zz", "bb", "cc", "ww", "ee", "ff", "gg", "hh", "ii", "jj", "kk",
           "ll", "mm", "nn" };
-    MyDataFrame         df;
+    MyDataFrame        df;
 
     df.load_data(std::move(idx),
                  std::make_pair("col_1", d1),
@@ -365,7 +365,8 @@ static void test_BiasVisitor()  {
         df.remove_data_by_loc({ 0, 1500 });
 
         using avg1 = MeanVisitor;
-        avg1                                avg1_v;
+
+        avg1                                     avg1_v;
         bias_v   bias1 (avg1_v);
 
         df.single_act_visit("IBM_Close", bias1);
@@ -381,7 +382,8 @@ static void test_BiasVisitor()  {
         assert(std::abs(bias1.get_result()[210] - 0.0242) < 0.0001);
 
         using s_avg1 = StableMeanVisitor;
-        s_avg1                              s_avg1_v;
+
+        s_avg1                                   s_avg1_v;
         bias_v s_bias1 (s_avg1_v);
 
         df.single_act_visit("IBM_Close", s_bias1);
@@ -397,7 +399,8 @@ static void test_BiasVisitor()  {
         assert(std::abs(s_bias1.get_result()[210] - 0.0242) < 0.0001);
 
         using avg2 = WeightedMeanVisitor;
-        avg2                                avg2_v;
+
+        avg2                                     avg2_v;
         bias_v   bias2 (avg2_v);
 
         df.single_act_visit("IBM_Close", bias2);
@@ -413,7 +416,8 @@ static void test_BiasVisitor()  {
         assert(std::abs(bias2.get_result()[210] - 0.0168) < 0.0001);
 
         using avg3 = GeometricMeanVisitor;
-        avg3                                avg3_v;
+
+        avg3                                     avg3_v;
         bias_v   bias3 (avg3_v);
 
         df.single_act_visit("IBM_Close", bias3);
@@ -429,7 +433,8 @@ static void test_BiasVisitor()  {
         assert(std::abs(bias3.get_result()[210] - 0.0245) < 0.0001);
 
         using avg4 = HarmonicMeanVisitor;
-        avg4                                avg4_v;
+
+        avg4                                     avg4_v;
         bias_v   bias4 (avg4_v);
 
         df.single_act_visit("IBM_Close", bias4);
@@ -1081,7 +1086,7 @@ static void test_TreynorRatioVisitor()  {
         { 0.2, 0.58, -0.60, -0.08, 0.05, 0.87, 0.2, 0.4, 0.5, 0.06, 0.3, -0.34,
           -0.9, 0.8, -0.4, 0.86, 0.01, 1.02, -0.02, -1.5, 0.2 };
     StlVecType            i1 = { 22, 23, 24, 25, 99 };
-    MyDataFrame                 df;
+    MyDataFrame                df;
 
     df.load_data(std::move(idx),
                  std::make_pair("asset", d1),
@@ -2174,14 +2179,12 @@ static void test_read_csv_with_vector()  {
 
         assert(df.get_index().size() == 564);
         assert((std::fabs(
-            df.get_column("Close")[4] - 1.0201) < 0.0001));
+                    df.get_column("Close")[4] - 1.0201) < 0.0001));
         assert((df.get_column("Volume")[4] == 3724291200));
-        assert((std::fabs(
-            df.get_column>
-                ("Z Score")[4][1] - -0.329) < 0.0001));
-        assert((std::fabs(
-            df.get_column>
-                ("Return Vector")[4][3] - -0.0182) < 0.0001));
+        assert((std::fabs(df.get_column>
+                              ("Z Score")[4][1] - -0.329) < 0.0001));
+        assert((std::fabs(df.get_column>
+                              ("Return Vector")[4][3] - -0.0182) < 0.0001));
         assert((std::isnan(
             df.get_column>("Return Vector")[10][0])));
     }
@@ -2211,11 +2214,13 @@ static void test_read_csv_with_maps()  {
         // df.write
         //     (std::cout, io_format::csv2);
         assert(df.get_index().size() == 4);
-        assert((std::fabs(df.get_column("Close")[3] - 1.0234) < 0.0001));
+        assert((std::fabs(df.get_column("Close")[3] - 1.0234) <
+                              0.0001));
         assert((df.get_column("Volume")[3] == 3605190400));
 
         assert((std::fabs(
-            df.get_column("Map 1")[3]["label four 2"] - -782.5) < 0.001));
+            df.get_column("Map 1")[3]["label four 2"] - -782.5) <
+                0.001));
         assert((std::fabs(
             df.get_column("Map 1")[0]["label one 1"] - 123.0) < 0.001));
         assert((std::fabs(
@@ -2479,6 +2484,287 @@ static void test_get_str_col_stats()  {
 
 // -----------------------------------------------------------------------------
 
+static void test_inversion_count()  {
+
+    std::cout << "\nTesting inversion_count(  ) ..." << std::endl;
+
+    using IntDataFrame = StdDataFrame;
+
+    std::vector    idx =
+        { 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17 };
+    std::vector    i1 =
+        { 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17 };
+    std::vector    i2 =
+        { 17, 16, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0 };
+    std::vector    i3 =
+        { 1, 0, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17 };
+    std::vector    i4 =
+        { 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 17, 16 };
+    std::vector    i5 =
+        { 1, 0, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 17, 16 };
+    std::vector    i6 =
+        { 17, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 0 };
+    std::vector    i7 = 
+        { 0, 1, 2, 3, 4, 5, 6, 10, 8, 9, 7, 11, 12, 13, 14, 15, 16, 17 };
+    std::vector    i8 =
+        { 0, 1, 2, 15, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 3, 16, 17 };
+    std::vector    i9 =
+        { 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2 };
+    std::vector    i10 = 
+        { 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 3 };
+    std::vector    i11 =
+        { 2, 2, 2, 2, 3, 2, 2, 2, 2, 4, 2, 2, 2, 5, 2, 2, 2, 6 };
+    IntDataFrame        df;
+
+    df.load_data(std::move(idx),
+                 std::make_pair("i1", i1),
+                 std::make_pair("i2", i2),
+                 std::make_pair("i3", i3),
+                 std::make_pair("i4", i4),
+                 std::make_pair("i5", i5),
+                 std::make_pair("i6", i6),
+                 std::make_pair("i7", i7),
+                 std::make_pair("i8", i8),
+                 std::make_pair("i9", i9),
+                 std::make_pair("i10", i10),
+                 std::make_pair("i11", i11));
+
+    assert(df.inversion_count("i1") == 0);
+    assert(df.inversion_count("i2") == 153);
+    assert(df.inversion_count("i3") == 1);
+    assert(df.inversion_count("i4") == 1);
+    assert(df.inversion_count("i5") == 2);
+    assert(df.inversion_count("i6") == 33);
+    assert(df.inversion_count("i7") == 5);
+    assert(df.inversion_count("i8") == 23);
+    assert(df.inversion_count("i9") == 153);
+    assert(df.inversion_count("i10") == 136);
+    assert(df.inversion_count("i11") == 110);
+
+    assert((df.inversion_count>("i1") == 153));
+    assert((df.inversion_count>("i2") == 0));
+}
+
+// -----------------------------------------------------------------------------
+
+static void test__like_clause_compare_()  {
+
+    std::cout << "\nTesting _like_clause_compare_(  ) ..." << std::endl;
+
+    const std::string   str1("345&%$abcM");
+    const std::string   str2("!@#$0987^HGTtiff\"");
+    const std::string   str3("ABFDTiy");
+
+    assert(_like_clause_compare_("345*", str1.c_str()));
+    assert(_like_clause_compare_("345*M", str1.c_str()));
+    assert(_like_clause_compare_("345*m", str1.c_str()) == false);
+    assert(_like_clause_compare_("345*m", str1.c_str(), true));
+    assert(_like_clause_compare_("?*[0-9][0-9][0-9][0-9]?*", str2.c_str()));
+    assert((_like_clause_compare_("?*[0-9][0-9][0-9][0-9][0-9]?*",
+                                  str2.c_str()) == false));
+    assert(_like_clause_compare_("?*\"", str2.c_str()));
+    assert(_like_clause_compare_("?*[^ABFDTiy]?*", str3.c_str()) == false);
+    assert(_like_clause_compare_("*[^WdrhID]*", str3.c_str()));
+}
+
+// -----------------------------------------------------------------------------
+
+static void test_get_data_by_like()  {
+
+    std::cout << "\nTesting get_data_by_like( ) ..." << std::endl;
+
+    StlVecType  idxvec =
+        { 1UL, 2UL, 3UL, 10UL, 5UL, 7UL, 8UL, 12UL, 9UL, 12UL, 10UL, 13UL,
+          10UL, 15UL, 14UL };
+    StlVecType    strvec1 =
+        { "345&%$abcM", "!@#$0987^HGTtiff\"", "ABFDTiy", "345&%$abcM",
+          "!@#$0987^HGTtiff\"", "!@#$0987^HGTtiff\"", "345&%$abcM",
+          "!@#$0987^HGTtiff\"", "ABFDTiy", "345&%$abcM", "!@#$0987^HGTtiff\"",
+          "ABFDTiy", "345&%$abcM", "ABFDTiy", "ABFDTiy" };
+    StlVecType    strvec2 =
+        { "ABFDTiy", "!@#$0987^HGTtiff\"", "ABFDTiy", "345&%$abcM",
+          "!@#$0987^HGTtiff\"", "ABFDTiy", "!@#$0987^HGTtiff\"",
+          "!@#$0987^HGTtiff\"", "ABFDTiy", "345&%$abcM", "!@#$0987^HGTtiff\"",
+          "ABFDTiy", "345&%$abcM", "ABFDTiy", "ABFDTiy" };
+    StlVecType            intvec =
+        { 1, 2, 3, 10, 5, 7, 8, 12, 9, 12, 10, 13, 10, 15, 14 };
+    MyDataFrame                df;
+
+    df.load_data(std::move(idxvec),
+                 std::make_pair("str column 1", strvec1),
+                 std::make_pair("str column 2", strvec2),
+                 std::make_pair("int column", intvec));
+
+    auto    df_like2 =
+        df.get_data_by_like(
+            "str column 1",
+            "str column 2",
+            "?*[0-9][0-9][0-9][0-9]?*",
+            "?*[0-9][0-9][0-9][0-9]?*");
+
+    assert(df_like2.get_index().size() == 4);
+    assert(df_like2.get_index()[2] == 12);
+    assert(df_like2.get_column("int column")[2] == 12);
+    assert(df_like2.get_column("str column 1").size() == 4);
+    assert(df_like2.get_column("str column 2").size() == 4);
+    assert((df_like2.get_column("str column 1")[0] ==
+                "!@#$0987^HGTtiff\""));
+    assert((df_like2.get_column("str column 1")[2] ==
+                "!@#$0987^HGTtiff\""));
+    assert((df_like2.get_column("str column 2")[0] ==
+                "!@#$0987^HGTtiff\""));
+    assert((df_like2.get_column("str column 2")[2] ==
+                "!@#$0987^HGTtiff\""));
+
+    auto    dfv_like2 =
+        df.get_view_by_like(
+            "str column 1",
+            "str column 2",
+            "?*[0-9][0-9][0-9][0-9]?*",
+            "?*[0-9][0-9][0-9][0-9]?*");
+
+    assert(dfv_like2.get_index().size() == 4);
+    assert(dfv_like2.get_index()[2] == 12);
+    assert(dfv_like2.get_column("int column")[2] == 12);
+    assert(dfv_like2.get_column("str column 1").size() == 4);
+    assert(dfv_like2.get_column("str column 2").size() == 4);
+    assert((dfv_like2.get_column("str column 1")[0] ==
+                "!@#$0987^HGTtiff\""));
+    assert((dfv_like2.get_column("str column 1")[2] ==
+                "!@#$0987^HGTtiff\""));
+    assert((dfv_like2.get_column("str column 2")[0] ==
+                "!@#$0987^HGTtiff\""));
+    assert((dfv_like2.get_column("str column 2")[2] ==
+                "!@#$0987^HGTtiff\""));
+
+    dfv_like2.get_column("str column 2")[3] = "ABC";
+    assert(dfv_like2.get_column("str column 2")[3] == "ABC");
+    assert(df.get_column("str column 2")[10] == "ABC");
+
+    auto    df_like1 =
+        df.get_data_by_like(
+            "str column 1",
+            "?*&%?*");
+
+    assert(df_like1.get_index().size() == 5);
+    assert(df_like1.get_index()[2] == 8);
+    assert(df_like1.get_column("int column")[2] == 8);
+    assert(df_like1.get_column("str column 1").size() == 5);
+    assert(df_like1.get_column("str column 2").size() == 5);
+    assert((df_like1.get_column("str column 1")[0] ==
+                "345&%$abcM"));
+    assert((df_like1.get_column("str column 1")[2] ==
+                "345&%$abcM"));
+    assert((df_like1.get_column("str column 2")[0] == "ABFDTiy"));
+    assert((df_like1.get_column("str column 2")[2] ==
+                "!@#$0987^HGTtiff\""));
+}
+
+// -----------------------------------------------------------------------------
+
+static void test_clear()  {
+
+    std::cout << "\nTesting clear( ) ..." << std::endl;
+
+    StlVecType  idx =
+        { 123450, 123451, 123452, 123453, 123454, 123455, 123456,
+          123457, 123458, 123459, 123460, 123461, 123462, 123466 };
+    StlVecType d1 = { 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14 };
+    StlVecType d2 = { 8, 9, 10, 11, 12, 13, 14, 20, 22, 23,
+                              30, 31, 32, 1.89 };
+    StlVecType d3 = { 15, 16, 17, 18, 19, 20, 21,
+                              0.34, 1.56, 0.34, 2.3, 0.1, 0.89, 0.45 };
+    StlVecType    i1 = { 22, 23, 24, 25, 99, 100, 101, 3, 2 };
+    StlVecType    strvec =
+        { "zz", "bb", "cc", "ww", "ee", "ff", "gg", "hh", "ii", "jj", "kk",
+          "ll", "mm", "nn" };
+    MyDataFrame        df1;
+
+    df1.load_data(std::move(idx),
+                  std::make_pair("col_1", d1),
+                  std::make_pair("col_2", d2),
+                  std::make_pair("col_3", d3),
+                  std::make_pair("col_4", i1),
+                  std::make_pair("str_col", strvec));
+
+    StlVecType  idx2 =
+        { 123450, 123451, 123452, 123453, 123454, 123455, 123456,
+          123457, 123458, 123459, 123460, 123461, 123462, 123466 };
+    StlVecType d12 = { 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14 };
+    StlVecType d22 = { 8, 9, 10, 11, 12, 13, 14, 20, 22, 23,
+                               30, 31, 32, 1.89 };
+    StlVecType d32 = { 15, 16, 17, 18, 19, 20, 21,
+                               0.34, 1.56, 0.34, 2.3, 0.1, 0.89, 0.45 };
+    StlVecType    i12 = { 22, 23, 24, 25, 99, 100, 101, 3, 2 };
+    StlVecType    strvec2 =
+        { "zz", "bb", "cc", "ww", "ee", "ff", "gg", "hh", "ii", "jj", "kk",
+          "ll", "mm", "nn" };
+    MyDataFrame        df2;
+
+    df2.load_data(std::move(idx2),
+                  std::make_pair("col_1", d12),
+                  std::make_pair("col_2", d22),
+                  std::make_pair("col_3", d32),
+                  std::make_pair("col_4", i12),
+                  std::make_pair("str_col", strvec2));
+
+    df1.clear();
+    assert(df1.empty());
+    assert(df1.shapeless());
+    assert(df2.get_index()[4] == 123454);
+    assert(df2.get_column("col_4")[7] == 3);
+    assert(df2.get_column("str_col")[5] == "ff");
+
+    df1 = df2;
+    assert(df1.get_index()[4] == 123454);
+    assert(df1.get_column("col_4")[7] == 3);
+    assert(df1.get_column("str_col")[5] == "ff");
+}
+
+// -----------------------------------------------------------------------------
+
+static void test_swap()  {
+
+    std::cout << "\nTesting swap( ) ..." << std::endl;
+
+    StlVecType  idx =
+        { 123450, 123451, 123452, 123453, 123454, 123455, 123456,
+          123457, 123458, 123459, 123460, 123461, 123462, 123466 };
+    StlVecType d1 = { 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14 };
+    StlVecType d2 = { 8, 9, 10, 11, 12, 13, 14, 20, 22, 23,
+                              30, 31, 32, 1.89 };
+    StlVecType d3 = { 15, 16, 17, 18, 19, 20, 21,
+                              0.34, 1.56, 0.34, 2.3, 0.1, 0.89, 0.45 };
+    StlVecType    i1 = { 22, 23, 24, 25, 99, 100, 101, 3, 2 };
+    StlVecType    strvec =
+        { "zz", "bb", "cc", "ww", "ee", "ff", "gg", "hh", "ii", "jj", "kk",
+          "ll", "mm", "nn" };
+    MyDataFrame        df1;
+    MyDataFrame        df2;
+
+    df1.load_data(std::move(idx),
+                  std::make_pair("col_1", d1),
+                  std::make_pair("col_2", d2),
+                  std::make_pair("col_3", d3),
+                  std::make_pair("col_4", i1),
+                  std::make_pair("str_col", strvec));
+
+    assert(df2.empty());
+    assert(df2.shapeless());
+    assert(df1.get_index()[4] == 123454);
+    assert(df1.get_column("col_4")[7] == 3);
+    assert(df1.get_column("str_col")[5] == "ff");
+
+    df1.swap(df2);
+    assert(df1.empty());
+    assert(df1.shapeless());
+    assert(df2.get_index()[4] == 123454);
+    assert(df2.get_column("col_4")[7] == 3);
+    assert(df2.get_column("str_col")[5] == "ff");
+}
+
+// -----------------------------------------------------------------------------
+
 int main(int, char *[]) {
 
     MyDataFrame::set_optimum_thread_level();
@@ -2533,6 +2819,11 @@ int main(int, char *[]) {
     test_PriceVolumeTrendVisitor();
     test_QuantQualEstimationVisitor();
     test_get_str_col_stats();
+    test_inversion_count();
+    test__like_clause_compare_();
+    test_get_data_by_like();
+    test_clear();
+    test_swap();
 
     return (0);
 }
diff --git a/test/dataframe_tester_output.txt b/test/dataframe_tester_output.txt
index 4c2b5e11..059df055 100644
--- a/test/dataframe_tester_output.txt
+++ b/test/dataframe_tester_output.txt
@@ -85,6 +85,12 @@ str_col:28::XXXX10,XXXX11,XXXX01,XXXX02,XXXX03,XXXX6,XXXX7,Running fast,
 dbl_col:28::2.009,3.111,10,4.2222,5.3333,12,6.25,10,0.9999,1.2345,4.2345,3,8,3.3333,2.2345,4.25,3.2345,0.009,1.111,5.25,11,5.2345,2.2222,1.009,2.111,9,3.2222,4.3333,
 dbl_col_2:28::0.87865,-0.6999,0.4111,0.1902,-0.4888,0.2,0.1056,0.1,0.06743,0.998,0.15678,0.923,0.0111,-0.8888,0.3456,0.0056,0.056,0.07865,-0.9999,0.0456,0.14,0.00345,0.1002,0.078654,-0.8999,0.01119,0.8002,-0.9888,
 bool_col:28::0,1,1,1,0,0,1,0,0,0,1,0,1,0,0,0,0,0,0,0,1,0,1,0,0,1,0,0,
+Map 1:4::3{label one 1:123|label one 2:-782.5|label one 3:444.44},3{label two 1:123|label two 2:-782.5|label two 3:444.44},3{label three 1:123|label three 2:-782.5|label three 3:444.44},3{label four 1:123|label four 2:-782.5|label four 3:444.44},
+Unordered Map:4::3{Key one 3:444.44|Key one 2:-782.5|Key one 1:123},3{Key two 3:444.44|Key two 2:-782.5|Key two 1:123},3{Key three 3:444.44|Key three 2:-782.5|Key three 1:123},3{Key four 2:-782.5|Key four 3:444.44|Key four 1:123},
+Str Vec:4::4[bbb|aaa|zzz|ddd],4[aaa|bbb|ccc|www],4[123|abc|345|list],3[bbb|aaa|zzz],
+Double Set:4::3[-782.5|123|444.44],3[-782.5|0|1],3[-782.5|123|444.44],4[-782.5|100.5|123|444.44],
+Str Set:4::3[-782.5|123.0|444.44],3[-782.5|1:123.0|:444.44],3[-782.5|123.0|444.44],4[-782.5|100.5|123.0|444.44],
+Z Score:4::10[1.95474040557|0.552535091086|0.775388936446|-0.561817339812|0.106794118727|-0.153218675013|-0.896114748672|-1.72258101434|-0.301804546072|0.246077772077],10[-0.985180680575|-0.338649566179|1.37000434149|0.831246802651|-0.415610988193|1.06213106869|0.554158098662|0.507981245453|-1.55472278822|-1.03135753378],10[-1.39575784008|-1.62506351709|-0.907239380237|-0.159508137551|0.807541881212|0.937157205458|0.578236204203|0.717820563726|-0.0398797142361|1.0866927346],10[1.94246107491|-0.062340594565|0.246115232403|-1.24462409799|-0.190844664632|0.8115331407|0.381021476571|-1.60448155299|-0.422151990754|0.143311976349],
 
 INDEX:28::string_index_1,string_index_2,string_index_3,string_index_4,string_index_5,string_index_6,string_index_7,string_index_8,string_index_9,string_index_10,string_index_11,string_index_12,string_index_13,string_index_14,string_index_15,string_index_16,string_index_17,string_index_18,string_index_19,string_index_20,string_index_21,string_index_22,string_index_23,string_index_24,string_index_25,string_index_26,string_index_27,string_index_28,
 ul_col:28::123450,123451,123452,123450,123455,123450,123449,123448,123451,123452,123452,123450,123455,123450,123454,123453,123456,123457,123458,123459,123460,123441,123442,123432,123433,123434,123435,123436,
@@ -93,6 +99,14 @@ str_col:28::XXXX10,XXXX11,XXXX01,XXXX02,XXXX03,XXXX6,XXXX7,Running fast,
 dbl_col:28::2.009,3.111,10,4.2222,5.3333,12,6.25,10,0.9999,1.2345,4.2345,3,8,3.3333,2.2345,4.25,3.2345,0.009,1.111,5.25,11,5.2345,2.2222,1.009,2.111,9,3.2222,4.3333,
 dbl_col_2:28::0.87865,-0.6999,0.4111,0.1902,-0.4888,0.2,0.1056,0.1,0.06743,0.998,0.15678,0.923,0.0111,-0.8888,0.3456,0.0056,0.056,0.07865,-0.9999,0.0456,0.14,0.00345,0.1002,0.078654,-0.8999,0.01119,0.8002,-0.9888,
 bool_col:28::0,1,1,1,0,0,1,0,0,0,1,0,1,0,0,0,0,0,0,0,1,0,1,0,0,1,0,0,
+char_col:6::C,%,x,0,A,!,
+uchar_col:6::C,%,250,0,A,!,
+Map 1:4::3{label one 1:123|label one 2:-782.5|label one 3:444.44},3{label two 1:123|label two 2:-782.5|label two 3:444.44},3{label three 1:123|label three 2:-782.5|label three 3:444.44},3{label four 1:123|label four 2:-782.5|label four 3:444.44},
+Unordered Map:4::3{Key one 3:444.44|Key one 2:-782.5|Key one 1:123},3{Key two 3:444.44|Key two 2:-782.5|Key two 1:123},3{Key three 3:444.44|Key three 2:-782.5|Key three 1:123},3{Key four 2:-782.5|Key four 3:444.44|Key four 1:123},
+Str Vec:4::4[bbb|aaa|zzz|ddd],4[aaa|bbb|ccc|www],4[123|abc|345|list],3[bbb|aaa|zzz],
+Double Set:4::3[-782.5|123|444.44],3[-782.5|0|1],3[-782.5|123|444.44],4[-782.5|100.5|123|444.44],
+Str Set:4::3[-782.5|123.0|444.44],3[-782.5|1:123.0|:444.44],3[-782.5|123.0|444.44],4[-782.5|100.5|123.0|444.44],
+Z Score:4::10[1.95474040557|0.552535091086|0.775388936446|-0.561817339812|0.106794118727|-0.153218675013|-0.896114748672|-1.72258101434|-0.301804546072|0.246077772077],10[-0.985180680575|-0.338649566179|1.37000434149|0.831246802651|-0.415610988193|1.06213106869|0.554158098662|0.507981245453|-1.55472278822|-1.03135753378],10[-1.39575784008|-1.62506351709|-0.907239380237|-0.159508137551|0.807541881212|0.937157205458|0.578236204203|0.717820563726|-0.0398797142361|1.0866927346],10[1.94246107491|-0.062340594565|0.246115232403|-1.24462409799|-0.190844664632|0.8115331407|0.381021476571|-1.60448155299|-0.422151990754|0.143311976349],
 
 INDEX:28::1547825036.3,1516179600.874123908,1516093200.234,1516006800.234098,1515920400.2309,1515834000.89,1515747600.123456789,1515661200.12309,1515574800.4562387,1515488400.2345609,1515402000.78,1515315600.340987645,1515229200.309812765,1515142800.93451984,1515056400.671092346,1514970000.450137234,1514883600.91256923,1514797200.67,1514624400.4562,1514538000.5,1514451600.0,1514365200.896120945,1514278800.783452098,378205200000.561209834,409741200000.346,441277200000.340987,472899600.0,504435600.871234561,
 ul_col:28::123450,123451,123452,123450,123455,123450,123449,123448,123451,123452,123452,123450,123455,123450,123454,123453,123456,123457,123458,123459,123460,123441,123442,123432,123433,123434,123435,123436,
@@ -751,12 +765,12 @@ col_3:1::15,
 col_str:1::11,
 col_4:1::22,
 
-INDEX:1::123451,
-col_1:1::2,
-col_2:1::9,
-col_3:1::16,
-col_str:1::22,
-col_4:1::23,
+INDEX:1::123452,
+col_1:1::3,
+col_2:1::10,
+col_3:1::17,
+col_str:1::33,
+col_4:1::24,
 
 
 Testing write(json) ...
@@ -1141,35 +1155,35 @@ INDEX:28:,ul_col:28:,xint_col:28:,str_col:25:,dbl_col
 123435,123435,45,,5.3333,,
 123450,123450,2,,,,
 
-INDEX:28:,ul_col:28:,dbl_col_2:26:,bool_col:6:,str_col:28:,xint_col:28:,dbl_col:27:
-123450,123450,0.998,1,4% of something,1,1.2345
-123451,123451,0.3456,1,Description 4/5,2,2.2345
-123452,123452,0.056,1,This is bad,3,3.2345
-123450,123450,0.15678,0,3.4% of GDP,4,4.2345
-123455,123455,0.00345,0,Market drops,5,5.2345
-123450,123450,0.923,1,Market pulls back,3,3
-123449,123449,0.06743,,$15 increase,7,0.9999
-123450,123450,0.1,,Running fast,3,10
-123451,123451,0.0056,,C++14 development,9,4.25
-123450,123450,0.07865,,Some explanation,10,0.009
-123452,123452,0.0111,,More strings,3,8
-123450,123450,0.1002,,Bonds vs. Equities,2,2.2222
-123455,123455,-0.8888,,Almost done,3,3.3333
-123450,123450,0.14,,XXXX04,14,11
-123454,123454,0.0456,,XXXX2,2,5.25
-123450,123450,0.078654,,XXXX3,2,1.009
-123450,123450,-0.8999,,XXXX4,2,2.111
-123457,123457,0.8002,,XXXX4,3,9
-123458,123458,-0.9888,,XXXX5,2,3.2222
-123459,123459,0.2,,XXXX6,3,4.3333
-123450,123450,0.1056,,XXXX7,3,12
-123441,123441,0.87865,,XXXX10,3,6.25
-123442,123442,-0.6999,,XXXX11,3,2.009
-123432,123432,0.4111,,XXXX02,3,3.111
-123450,123450,0.1902,,XXXX03,36,10
-123450,123450,-0.4888,,,2,4.2222
-123435,123435,,,,45,5.3333
-123450,123450,,,,2,
+INDEX:28:,ul_col:28:,dbl_col_2:26:,bool_col:6:,str_col:28:,xint_col:28:,dbl_col:27:,,
+123450,123450,0.998,1,4% of something,1,1.2345,,
+123451,123451,0.3456,1,Description 4/5,2,2.2345,,
+123452,123452,0.056,1,This is bad,3,3.2345,,
+123450,123450,0.15678,0,3.4% of GDP,4,4.2345,,
+123455,123455,0.00345,0,Market drops,5,5.2345,,
+123450,123450,0.923,1,Market pulls back,3,3,,
+123449,123449,0.06743,,$15 increase,7,0.9999,,
+123450,123450,0.1,,Running fast,3,10,,
+123451,123451,0.0056,,C++14 development,9,4.25,,
+123450,123450,0.07865,,Some explanation,10,0.009,,
+123452,123452,0.0111,,More strings,3,8,,
+123450,123450,0.1002,,Bonds vs. Equities,2,2.2222,,
+123455,123455,-0.8888,,Almost done,3,3.3333,,
+123450,123450,0.14,,XXXX04,14,11,,
+123454,123454,0.0456,,XXXX2,2,5.25,,
+123450,123450,0.078654,,XXXX3,2,1.009,,
+123450,123450,-0.8999,,XXXX4,2,2.111,,
+123457,123457,0.8002,,XXXX4,3,9,,
+123458,123458,-0.9888,,XXXX5,2,3.2222,,
+123459,123459,0.2,,XXXX6,3,4.3333,,
+123450,123450,0.1056,,XXXX7,3,12,,
+123441,123441,0.87865,,XXXX10,3,6.25,,
+123442,123442,-0.6999,,XXXX11,3,2.009,,
+123432,123432,0.4111,,XXXX02,3,3.111,,
+123450,123450,0.1902,,XXXX03,36,10,,
+123450,123450,-0.4888,,,2,4.2222,,
+123435,123435,,,,45,5.3333,,
+123450,123450,,,,2,,,
 
 
 Testing BoxCoxVisitor{ } ...
@@ -1237,6 +1251,7 @@ Testing GarmanKlassVolVisitor{  } ...
 
 Testing YangZhangVolVisitor{  } ...
 
+
 Testing no_index_writes ...
 INDEX:28::123450,123451,123452,123450,123455,123450,123449,123450,123451,123450,123452,123450,123455,123450,123454,123450,123450,123457,123458,123459,123450,123441,123442,123432,123450,123450,123435,123450,
 ul_col:28::123450,123451,123452,123450,123455,123450,123449,123450,123451,123450,123452,123450,123455,123450,123454,123450,123450,123457,123458,123459,123450,123441,123442,123432,123450,123450,123435,123450,
@@ -1339,83 +1354,83 @@ ul_col:28:,xint_col:28:,str_col:25:,dbl_col:27:,dbl_
 }
 
 Testing no_index_reads ...
-INDEX:28:,ul_col:28:,dbl_col_2:26:,bool_col:6:,str_col:28:,xint_col:28:,dbl_col:27:,ul_col_2:28:,dbl_col_2_2:26:,bool_col_2:6:,str_col_2:28:,xint_col_2:28:,dbl_col_3_2:27:,ul_col_no_idx:28:,dbl_col_2_no_idx:26:,bool_col_no_idx:6:,str_col_no_idx:28:,xint_col_no_idx:28:,dbl_col_no_idx:27:
-123450,123450,0.998,1,4% of something,1,1.2345,123450,0.998,1,4% of something,1,1.2345,123450,0.998,1,4% of something,1,1.2345
-123451,123451,0.3456,1,Description 4/5,2,2.2345,123451,0.3456,1,Description 4/5,2,2.2345,123451,0.3456,1,Description 4/5,2,2.2345
-123452,123452,0.056,1,This is bad,3,3.2345,123452,0.056,1,This is bad,3,3.2345,123452,0.056,1,This is bad,3,3.2345
-123450,123450,0.15678,0,3.4% of GDP,4,4.2345,123450,0.15678,0,3.4% of GDP,4,4.2345,123450,0.15678,0,3.4% of GDP,4,4.2345
-123455,123455,0.00345,0,Market drops,5,5.2345,123455,0.00345,0,Market drops,5,5.2345,123455,0.00345,0,Market drops,5,5.2345
-123450,123450,0.923,1,Market pulls back,3,3,123450,0.923,1,Market pulls back,3,3,123450,0.923,1,Market pulls back,3,3
-123449,123449,0.06743,,$15 increase,7,0.9999,123449,0.06743,,$15 increase,7,0.9999,123449,0.06743,,$15 increase,7,0.9999
-123450,123450,0.1,,Running fast,3,10,123450,0.1,,Running fast,3,10,123450,0.1,,Running fast,3,10
-123451,123451,0.0056,,C++14 development,9,4.25,123451,0.0056,,C++14 development,9,4.25,123451,0.0056,,C++14 development,9,4.25
-123450,123450,0.07865,,Some explanation,10,0.009,123450,0.07865,,Some explanation,10,0.009,123450,0.07865,,Some explanation,10,0.009
-123452,123452,0.0111,,More strings,3,8,123452,0.0111,,More strings,3,8,123452,0.0111,,More strings,3,8
-123450,123450,0.1002,,Bonds vs. Equities,2,2.2222,123450,0.1002,,Bonds vs. Equities,2,2.2222,123450,0.1002,,Bonds vs. Equities,2,2.2222
-123455,123455,-0.8888,,Almost done,3,3.3333,123455,-0.8888,,Almost done,3,3.3333,123455,-0.8888,,Almost done,3,3.3333
-123450,123450,0.14,,XXXX04,14,11,123450,0.14,,XXXX04,14,11,123450,0.14,,XXXX04,14,11
-123454,123454,0.0456,,XXXX2,2,5.25,123454,0.0456,,XXXX2,2,5.25,123454,0.0456,,XXXX2,2,5.25
-123450,123450,0.078654,,XXXX3,2,1.009,123450,0.078654,,XXXX3,2,1.009,123450,0.078654,,XXXX3,2,1.009
-123450,123450,-0.8999,,XXXX4,2,2.111,123450,-0.8999,,XXXX4,2,2.111,123450,-0.8999,,XXXX4,2,2.111
-123457,123457,0.8002,,XXXX4,3,9,123457,0.8002,,XXXX4,3,9,123457,0.8002,,XXXX4,3,9
-123458,123458,-0.9888,,XXXX5,2,3.2222,123458,-0.9888,,XXXX5,2,3.2222,123458,-0.9888,,XXXX5,2,3.2222
-123459,123459,0.2,,XXXX6,3,4.3333,123459,0.2,,XXXX6,3,4.3333,123459,0.2,,XXXX6,3,4.3333
-123450,123450,0.1056,,XXXX7,3,12,123450,0.1056,,XXXX7,3,12,123450,0.1056,,XXXX7,3,12
-123441,123441,0.87865,,XXXX10,3,6.25,123441,0.87865,,XXXX10,3,6.25,123441,0.87865,,XXXX10,3,6.25
-123442,123442,-0.6999,,XXXX11,3,2.009,123442,-0.6999,,XXXX11,3,2.009,123442,-0.6999,,XXXX11,3,2.009
-123432,123432,0.4111,,XXXX02,3,3.111,123432,0.4111,,XXXX02,3,3.111,123432,0.4111,,XXXX02,3,3.111
-123450,123450,0.1902,,XXXX03,36,10,123450,0.1902,,XXXX03,36,10,123450,0.1902,,XXXX03,36,10
-123450,123450,-0.4888,,,2,4.2222,123450,-0.4888,,,2,4.2222,123450,-0.4888,,,2,4.2222
-123435,123435,,,,45,5.3333,123435,,,,45,5.3333,123435,,,,45,5.3333
-123450,123450,,,,2,,123450,,,,2,,123450,,,,2,
-
-
-
-INDEX:28:,ul_col:28:,xint_col:28:,str_col:28:,dbl_col:28:,dbl_col_2:28:,bool_col:28:,ul_col_2:28:,xint_col_2:28:,str_col_2:28:,dbl_col_3:28:,dbl_col_2_2:28:,bool_col_2:28:,ul_col_no_idx:28:,xint_col_no_idx:28:,str_col_no_idx:28:,dbl_col_no_idx:28:,dbl_col_2_no_idx:28:,bool_col_no_idx:28:
-123432,123450,35,XXXX10,2.009,0.87865,0,123450,35,XXXX10,2.009,0.87865,0,123450,35,XXXX10,2.009,0.87865,0
-123433,123451,36,XXXX11,3.111,-0.6999,1,123451,36,XXXX11,3.111,-0.6999,1,123451,36,XXXX11,3.111,-0.6999,1
-123434,123452,40,XXXX01,10,0.4111,1,123452,40,XXXX01,10,0.4111,1,123452,40,XXXX01,10,0.4111,1
-123435,123450,45,XXXX02,4.2222,0.1902,1,123450,45,XXXX02,4.2222,0.1902,1,123450,45,XXXX02,4.2222,0.1902,1
-123436,123455,46,XXXX03,5.3333,-0.4888,0,123455,46,XXXX03,5.3333,-0.4888,0,123455,46,XXXX03,5.3333,-0.4888,0
-123441,123450,33,XXXX6,12,0.2,0,123450,33,XXXX6,12,0.2,0,123450,33,XXXX6,12,0.2,0
-123442,123449,34,XXXX7,6.25,0.1056,1,123449,34,XXXX7,6.25,0.1056,1,123449,34,XXXX7,6.25,0.1056,1
-123448,123448,8,Running fast,10,0.1,0,123448,8,Running fast,10,0.1,0,123448,8,Running fast,10,0.1,0
-123449,123451,7,$15 increase,0.9999,0.06743,0,123451,7,$15 increase,0.9999,0.06743,0,123451,7,$15 increase,0.9999,0.06743,0
-123450,123452,1,4% of something,1.2345,0.998,0,123452,1,4% of something,1.2345,0.998,0,123452,1,4% of something,1.2345,0.998,0
-123450,123452,4,3.4% of GDP,4.2345,0.15678,1,123452,4,3.4% of GDP,4.2345,0.15678,1,123452,4,3.4% of GDP,4.2345,0.15678,1
-123450,123450,6,Market pulls back,3,0.923,0,123450,6,Market pulls back,3,0.923,0,123450,6,Market pulls back,3,0.923,0
-123450,123455,12,Bonds vs. Equities,8,0.0111,1,123455,12,Bonds vs. Equities,8,0.0111,1,123455,12,Bonds vs. Equities,8,0.0111,1
-123450,123450,14,Here comes the sun,3.3333,-0.8888,0,123450,14,Here comes the sun,3.3333,-0.8888,0,123450,14,Here comes the sun,3.3333,-0.8888,0
-123451,123454,2,Description 4/5,2.2345,0.3456,0,123454,2,Description 4/5,2.2345,0.3456,0,123454,2,Description 4/5,2.2345,0.3456,0
-123451,123453,9,C++14 development,4.25,0.0056,0,123453,9,C++14 development,4.25,0.0056,0,123453,9,C++14 development,4.25,0.0056,0
-123452,123456,3,This is bad,3.2345,0.056,0,123456,3,This is bad,3.2345,0.056,0,123456,3,This is bad,3.2345,0.056,0
-123452,123457,10,Some explanation,0.009,0.07865,0,123457,10,Some explanation,0.009,0.07865,0,123457,10,Some explanation,0.009,0.07865,0
-123452,123458,11,More strings,1.111,-0.9999,0,123458,11,More strings,1.111,-0.9999,0,123458,11,More strings,1.111,-0.9999,0
-123453,123459,20,XXXX04,5.25,0.0456,0,123459,20,XXXX04,5.25,0.0456,0,123459,20,XXXX04,5.25,0.0456,0
-123454,123460,15,XXXX1,11,0.14,1,123460,15,XXXX1,11,0.14,1,123460,15,XXXX1,11,0.14,1
-123455,123441,5,Market drops,5.2345,0.00345,0,123441,5,Market drops,5.2345,0.00345,0,123441,5,Market drops,5.2345,0.00345,0
-123455,123442,13,Almost done,2.2222,0.1002,1,123442,13,Almost done,2.2222,0.1002,1,123442,13,Almost done,2.2222,0.1002,1
-123456,123432,22,XXXX2,1.009,0.078654,0,123432,22,XXXX2,1.009,0.078654,0,123432,22,XXXX2,1.009,0.078654,0
-123457,123433,23,XXXX3,2.111,-0.8999,0,123433,23,XXXX3,2.111,-0.8999,0,123433,23,XXXX3,2.111,-0.8999,0
-123458,123434,24,XXXX4,9,0.01119,1,123434,24,XXXX4,9,0.01119,1,123434,24,XXXX4,9,0.01119,1
-123459,123435,25,XXXX4,3.2222,0.8002,0,123435,25,XXXX4,3.2222,0.8002,0,123435,25,XXXX4,3.2222,0.8002,0
-123460,123436,30,XXXX5,4.3333,-0.9888,0,123436,30,XXXX5,4.3333,-0.9888,0,123436,30,XXXX5,4.3333,-0.9888,0
-
-
-
-INDEX:12:,col_3:12:,col_4:6:,col_str:12:,col_2:12:,col_1:12:,col_3_2:12:,col_4_2:6:,col_str_2:12:,col_2_2:12:,col_1_2:12:,col_3_no_idx:12:,col_4_no_idx:6:,col_str_no_idx:12:,col_2_no_idx:12:,col_1_no_idx:12:
-123450,15,22,11,8,1,15,22,11,8,1,15,22,11,8,1
-123451,16,23,22,9,2,16,23,22,9,2,16,23,22,9,2
-123452,17,24,33,10,3,17,24,33,10,3,17,24,33,10,3
-123453,18,25,aa,11,4,18,25,aa,11,4,18,25,aa,11,4
-123454,19,26,bb,12,5,19,26,bb,12,5,19,26,bb,12,5
-123455,20,27,cc,13,6,20,27,cc,13,6,20,27,cc,13,6
-123456,21,,dd,14,7,21,,dd,14,7,21,,dd,14,7
-123457,22,,tt,15,8,22,,tt,15,8,22,,tt,15,8
-123458,23,,uu,16,9,23,,uu,16,9,23,,uu,16,9
-123459,24,,ii,17,10,24,,ii,17,10,24,,ii,17,10
-123460,25,,88,18,11,25,,88,18,11,25,,88,18,11
-555555,555.543,,This is a test,777.78,55.55,555.543,,This is a test,777.78,55.55,555.543,,This is a test,777.78,55.55
+INDEX:28:,ul_col:28:,dbl_col_2:26:,bool_col:6:,str_col:28:,xint_col:28:,dbl_col:27:,char_col:25:,uchar_col:25:,ul_col_2:28:,dbl_col_2_2:26:,bool_col_2:6:,str_col_2:28:,xint_col_2:28:,dbl_col_3_2:27:,ul_col_no_idx:28:,dbl_col_2_no_idx:26:,bool_col_no_idx:6:,str_col_no_idx:28:,xint_col_no_idx:28:,dbl_col_no_idx:27:
+123450,123450,0.998,1,4% of something,1,1.2345,C,B,123450,0.998,1,4% of something,1,1.2345,123450,0.998,1,4% of something,1,1.2345
+123451,123451,0.3456,1,Description 4/5,2,2.2345,23,B,123451,0.3456,1,Description 4/5,2,2.2345,123451,0.3456,1,Description 4/5,2,2.2345
+123452,123452,0.056,1,This is bad,3,3.2345,^,&,123452,0.056,1,This is bad,3,3.2345,123452,0.056,1,This is bad,3,3.2345
+123450,123450,0.15678,0,3.4% of GDP,4,4.2345,F,250,123450,0.15678,0,3.4% of GDP,4,4.2345,123450,0.15678,0,3.4% of GDP,4,4.2345
+123455,123455,0.00345,0,Market drops,5,5.2345,x,B,123455,0.00345,0,Market drops,5,5.2345,123455,0.00345,0,Market drops,5,5.2345
+123450,123450,0.923,1,Market pulls back,3,3,N,B,123450,0.923,1,Market pulls back,3,3,123450,0.923,1,Market pulls back,3,3
+123449,123449,0.06743,,$15 increase,7,0.9999,C,B,123449,0.06743,,$15 increase,7,0.9999,123449,0.06743,,$15 increase,7,0.9999
+123450,123450,0.1,,Running fast,3,10,C,B,123450,0.1,,Running fast,3,10,123450,0.1,,Running fast,3,10
+123451,123451,0.0056,,C++14 development,9,4.25,C,B,123451,0.0056,,C++14 development,9,4.25,123451,0.0056,,C++14 development,9,4.25
+123450,123450,0.07865,,Some explanation,10,0.009,C,B,123450,0.07865,,Some explanation,10,0.009,123450,0.07865,,Some explanation,10,0.009
+123452,123452,0.0111,,More strings,3,8,C,(,123452,0.0111,,More strings,3,8,123452,0.0111,,More strings,3,8
+123450,123450,0.1002,,Bonds vs. Equities,2,2.2222,C,B,123450,0.1002,,Bonds vs. Equities,2,2.2222,123450,0.1002,,Bonds vs. Equities,2,2.2222
+123455,123455,-0.8888,,Almost done,3,3.3333,C,B,123455,-0.8888,,Almost done,3,3.3333,123455,-0.8888,,Almost done,3,3.3333
+123450,123450,0.14,,XXXX04,14,11,C,B,123450,0.14,,XXXX04,14,11,123450,0.14,,XXXX04,14,11
+123454,123454,0.0456,,XXXX2,2,5.25,C,B,123454,0.0456,,XXXX2,2,5.25,123454,0.0456,,XXXX2,2,5.25
+123450,123450,0.078654,,XXXX3,2,1.009,C,B,123450,0.078654,,XXXX3,2,1.009,123450,0.078654,,XXXX3,2,1.009
+123450,123450,-0.8999,,XXXX4,2,2.111,C,B,123450,-0.8999,,XXXX4,2,2.111,123450,-0.8999,,XXXX4,2,2.111
+123457,123457,0.8002,,XXXX4,3,9,C,B,123457,0.8002,,XXXX4,3,9,123457,0.8002,,XXXX4,3,9
+123458,123458,-0.9888,,XXXX5,2,3.2222,C,B,123458,-0.9888,,XXXX5,2,3.2222,123458,-0.9888,,XXXX5,2,3.2222
+123459,123459,0.2,,XXXX6,3,4.3333,C,B,123459,0.2,,XXXX6,3,4.3333,123459,0.2,,XXXX6,3,4.3333
+123450,123450,0.1056,,XXXX7,3,12,C,B,123450,0.1056,,XXXX7,3,12,123450,0.1056,,XXXX7,3,12
+123441,123441,0.87865,,XXXX10,3,6.25,C,B,123441,0.87865,,XXXX10,3,6.25,123441,0.87865,,XXXX10,3,6.25
+123442,123442,-0.6999,,XXXX11,3,2.009,C,B,123442,-0.6999,,XXXX11,3,2.009,123442,-0.6999,,XXXX11,3,2.009
+123432,123432,0.4111,,XXXX02,3,3.111,C,B,123432,0.4111,,XXXX02,3,3.111,123432,0.4111,,XXXX02,3,3.111
+123450,123450,0.1902,,XXXX03,36,10,C,B,123450,0.1902,,XXXX03,36,10,123450,0.1902,,XXXX03,36,10
+123450,123450,-0.4888,,,2,4.2222,,,123450,-0.4888,,,2,4.2222,123450,-0.4888,,,2,4.2222
+123435,123435,,,,45,5.3333,,,123435,,,,45,5.3333,123435,,,,45,5.3333
+123450,123450,,,,2,,,,123450,,,,2,,123450,,,,2,
+
+
+
+INDEX:28:,ul_col:28:,xint_col:28:,str_col:28:,dbl_col:28:,dbl_col_2:28:,bool_col:28:,Map 1:4:,Unordered Map:4:,Str Vec:4:,Double Set:4:,Str Set:4:,Z Score:4:,ul_col_2:28:,xint_col_2:28:,str_col_2:28:,dbl_col_3:28:,dbl_col_2_2:28:,bool_col_2:28:,ul_col_no_idx:28:,xint_col_no_idx:28:,str_col_no_idx:28:,dbl_col_no_idx:28:,dbl_col_2_no_idx:28:,bool_col_no_idx:28:,char_col:11:
+123432,123450,35,XXXX10,2.009,0.87865,0,3{label one 1:123|label one 2:-782.5|label one 3:444.44},3{Key one 3:444.44|Key one 2:-782.5|Key one 1:123},4[bbb|aaa|zzz|ddd],3[-782.5|123|444.44],3[-782.5|123.0|444.44],10[1.95474040557|0.552535091086|0.775388936446|-0.561817339812|0.106794118727|-0.153218675013|-0.896114748672|-1.72258101434|-0.301804546072|0.246077772077],123450,35,XXXX10,2.009,0.87865,0,123450,35,XXXX10,2.009,0.87865,0,C
+123433,123451,36,XXXX11,3.111,-0.6999,1,3{label two 1:123|label two 2:-782.5|label two 3:444.44},3{Key two 3:444.44|Key two 2:-782.5|Key two 1:123},4[aaa|bbb|ccc|www],3[-782.5|0|1],3[-782.5|1:123.0|:444.44],10[-0.985180680575|-0.338649566179|1.37000434149|0.831246802651|-0.415610988193|1.06213106869|0.554158098662|0.507981245453|-1.55472278822|-1.03135753378],123451,36,XXXX11,3.111,-0.6999,1,123451,36,XXXX11,3.111,-0.6999,1,%
+123434,123452,40,XXXX01,10,0.4111,1,3{label three 1:123|label three 2:-782.5|label three 3:444.44},3{Key three 3:444.44|Key three 2:-782.5|Key three 1:123},4[123|abc|345|list],3[-782.5|123|444.44],3[-782.5|123.0|444.44],10[-1.39575784008|-1.62506351709|-0.907239380237|-0.159508137551|0.807541881212|0.937157205458|0.578236204203|0.717820563726|-0.0398797142361|1.0866927346],123452,40,XXXX01,10,0.4111,1,123452,40,XXXX01,10,0.4111,1,x
+123435,123450,45,XXXX02,4.2222,0.1902,1,3{label four 1:123|label four 2:-782.5|label four 3:444.44},3{Key four 2:-782.5|Key four 3:444.44|Key four 1:123},3[bbb|aaa|zzz],4[-782.5|100.5|123|444.44],4[-782.5|100.5|123.0|444.44],10[1.94246107491|-0.062340594565|0.246115232403|-1.24462409799|-0.190844664632|0.8115331407|0.381021476571|-1.60448155299|-0.422151990754|0.143311976349],123450,45,XXXX02,4.2222,0.1902,1,123450,45,XXXX02,4.2222,0.1902,1,0
+123436,123455,46,XXXX03,5.3333,-0.4888,0,,,,,,,123455,46,XXXX03,5.3333,-0.4888,0,123455,46,XXXX03,5.3333,-0.4888,0,A
+123441,123450,33,XXXX6,12,0.2,0,,,,,,,123450,33,XXXX6,12,0.2,0,123450,33,XXXX6,12,0.2,0,0
+123442,123449,34,XXXX7,6.25,0.1056,1,,,,,,,123449,34,XXXX7,6.25,0.1056,1,123449,34,XXXX7,6.25,0.1056,1,%
+123448,123448,8,Running fast,10,0.1,0,,,,,,,123448,8,Running fast,10,0.1,0,123448,8,Running fast,10,0.1,0,-6
+123449,123451,7,$15 increase,0.9999,0.06743,0,,,,,,,123451,7,$15 increase,0.9999,0.06743,0,123451,7,$15 increase,0.9999,0.06743,0,0
+123450,123452,1,4% of something,1.2345,0.998,0,,,,,,,123452,1,4% of something,1.2345,0.998,0,123452,1,4% of something,1.2345,0.998,0,A
+123450,123452,4,3.4% of GDP,4.2345,0.15678,1,,,,,,,123452,4,3.4% of GDP,4.2345,0.15678,1,123452,4,3.4% of GDP,4.2345,0.15678,1,0
+123450,123450,6,Market pulls back,3,0.923,0,,,,,,,123450,6,Market pulls back,3,0.923,0,123450,6,Market pulls back,3,0.923,0,
+123450,123455,12,Bonds vs. Equities,8,0.0111,1,,,,,,,123455,12,Bonds vs. Equities,8,0.0111,1,123455,12,Bonds vs. Equities,8,0.0111,1,
+123450,123450,14,Here comes the sun,3.3333,-0.8888,0,,,,,,,123450,14,Here comes the sun,3.3333,-0.8888,0,123450,14,Here comes the sun,3.3333,-0.8888,0,
+123451,123454,2,Description 4/5,2.2345,0.3456,0,,,,,,,123454,2,Description 4/5,2.2345,0.3456,0,123454,2,Description 4/5,2.2345,0.3456,0,
+123451,123453,9,C++14 development,4.25,0.0056,0,,,,,,,123453,9,C++14 development,4.25,0.0056,0,123453,9,C++14 development,4.25,0.0056,0,
+123452,123456,3,This is bad,3.2345,0.056,0,,,,,,,123456,3,This is bad,3.2345,0.056,0,123456,3,This is bad,3.2345,0.056,0,
+123452,123457,10,Some explanation,0.009,0.07865,0,,,,,,,123457,10,Some explanation,0.009,0.07865,0,123457,10,Some explanation,0.009,0.07865,0,
+123452,123458,11,More strings,1.111,-0.9999,0,,,,,,,123458,11,More strings,1.111,-0.9999,0,123458,11,More strings,1.111,-0.9999,0,
+123453,123459,20,XXXX04,5.25,0.0456,0,,,,,,,123459,20,XXXX04,5.25,0.0456,0,123459,20,XXXX04,5.25,0.0456,0,
+123454,123460,15,XXXX1,11,0.14,1,,,,,,,123460,15,XXXX1,11,0.14,1,123460,15,XXXX1,11,0.14,1,
+123455,123441,5,Market drops,5.2345,0.00345,0,,,,,,,123441,5,Market drops,5.2345,0.00345,0,123441,5,Market drops,5.2345,0.00345,0,
+123455,123442,13,Almost done,2.2222,0.1002,1,,,,,,,123442,13,Almost done,2.2222,0.1002,1,123442,13,Almost done,2.2222,0.1002,1,
+123456,123432,22,XXXX2,1.009,0.078654,0,,,,,,,123432,22,XXXX2,1.009,0.078654,0,123432,22,XXXX2,1.009,0.078654,0,
+123457,123433,23,XXXX3,2.111,-0.8999,0,,,,,,,123433,23,XXXX3,2.111,-0.8999,0,123433,23,XXXX3,2.111,-0.8999,0,
+123458,123434,24,XXXX4,9,0.01119,1,,,,,,,123434,24,XXXX4,9,0.01119,1,123434,24,XXXX4,9,0.01119,1,
+123459,123435,25,XXXX4,3.2222,0.8002,0,,,,,,,123435,25,XXXX4,3.2222,0.8002,0,123435,25,XXXX4,3.2222,0.8002,0,
+123460,123436,30,XXXX5,4.3333,-0.9888,0,,,,,,,123436,30,XXXX5,4.3333,-0.9888,0,123436,30,XXXX5,4.3333,-0.9888,0,
+
+
+
+INDEX:12:,col_3:12:,col_4:6:,col_str:12:,col_2:12:,col_1:12:,col_char:12:,col_uchar:12:,col_3_2:12:,col_4_2:6:,col_str_2:12:,col_2_2:12:,col_1_2:12:,col_3_no_idx:12:,col_4_no_idx:6:,col_str_no_idx:12:,col_2_no_idx:12:,col_1_no_idx:12:,col_char_no_idx:12:,col_uchar_no_idx:12:
+123450,15,22,11,8,1,8,8,15,22,11,8,1,15,22,11,8,1,8,8
+123451,16,23,22,9,2,C,C,16,23,22,9,2,16,23,22,9,2,C,C
+123452,17,24,33,10,3,F,F,17,24,33,10,3,17,24,33,10,3,F,F
+123453,18,25,aa,11,4,$,$,18,25,aa,11,4,18,25,aa,11,4,$,$
+123454,19,26,bb,12,5,0,0,19,26,bb,12,5,19,26,bb,12,5,0,0
+123455,20,27,cc,13,6,8,8,20,27,cc,13,6,20,27,cc,13,6,8,8
+123456,21,,dd,14,7,x,220,21,,dd,14,7,21,,dd,14,7,x,220
+123457,22,,tt,15,8,X,X,22,,tt,15,8,22,,tt,15,8,X,X
+123458,23,,uu,16,9,h,h,23,,uu,16,9,23,,uu,16,9,h,h
+123459,24,,ii,17,10,u,&,24,,ii,17,10,24,,ii,17,10,u,&
+123460,25,,88,18,11,18,18,25,,88,18,11,25,,88,18,11,18,18
+555555,555.543,,This is a test,777.78,55.55,N,255,555.543,,This is a test,777.78,55.55,555.543,,This is a test,777.78,55.55,N,255
 
 
 Testing KamaVisitor{  } ...
@@ -1999,6 +2014,12 @@ Testing PriceVolumeTrendVisitor{  } ...
 Testing QuantQualEstimationVisitor{  } ...
 
 Testing get_str_col_stats(  ) ...
+
+Testing inversion_count(  ) ...
+
+Testing _like_clause_compare_(  ) ...
+
+Testing get_data_by_like( ) ...
 Hello World!
 Str Column = A, B, C, D, E, F, G, H, I, J, 
 There are 5031 IBM close prices
@@ -2018,7 +2039,7 @@ INDEX:10:,string col:10:,Cool Column:10:,numbers:10: