Merge branch 'release/0.9.0'

DESCRIPTION

@@ -1,8 +1,8 @@
 Package: tableone
 Type: Package
 Title: Create 'Table 1' to Describe Baseline Characteristics
-Version: 0.8.1
-Date: 2017-06-16
+Version: 0.9.0
+Date: 2017-11-03
 Author: Kazuki Yoshida, Justin Bohn.
 Maintainer: Kazuki Yoshida <[email protected]>
 Description: Creates 'Table 1', i.e., description of baseline patient
@@ -20,7 +20,9 @@ Imports:
     e1071,
     zoo,
     gmodels,
-    nlme
+    nlme,
+    lmerTest,
+    labelled
 Suggests:
     survival,
     testthat,

NEWS

@@ -1,3 +1,33 @@
+tableone 0.9.0 (2017-11-03)
+----------------------------------------------------------------
+
+NEW FEATURES
+
+* The "varLabels" option for the print.TableOne method was added.
+  When TRUE, instead of printing the variable names, their
+  corresponding variable labels are used. Variable labels must be
+  stored in the data frame to be used via labelled::var_label
+  function. This option is also available in ExtractSmd function.
+
+* The "dropEqual" option for the print methods was implemented. If
+  TRUE, the level description for two-level variables such as " = 1"
+  and " = TRUE" are not shown. This can obscure what level is being
+  shown depending on the variable naming scheme, thus, should only
+  be used after the initial results were checked for correctness.
+
+BUG FIXES
+
+* Handling of lme4 models was improved in ShowRegTable.
+
+
+tableone 0.8.2 (2017-08-31)
+----------------------------------------------------------------
+
+BUG FIXES
+
+* Handle labelled class as categorical.
+
+
 tableone 0.8.1 (2017-06-17)
 ----------------------------------------------------------------
 

R/CreateTableOne.R

@@ -129,6 +129,10 @@ function(vars,                                      # character vector of variab
     ## Abort if no variables exist at this point
     ModuleStopIfNoVarsLeft(vars)
 
+    ## Get variable labels (named list of label string or NULL)
+    ## Need to occur before applying factor().
+    varLabels <- labelled::var_label(data[vars])
+
     ## Factor conversions if the factorVars argument exist
     if (!missing(factorVars)) {
         ## Check if variables exist. Drop them if not.
@@ -149,7 +153,7 @@ function(vars,                                      # character vector of variab
 
     ## Classify as varFactors if any one of these classes are contained
     varFactors <-sapply(varClasses, function(VEC) {
-        any(VEC %in% c("factor", "ordered", "logical", "character"))
+        any(VEC %in% c("factor", "ordered", "logical", "character", "labelled"))
     })
     varFactors <- names(varFactors)[varFactors]
 
@@ -243,7 +247,9 @@ function(vars,                                      # character vector of variab
                                             varFactors  = varFactors,
                                             varNumerics = varNumerics,
                                             ## Missing data percentage for each variable (no strata).
-                                            percentMissing = percentMissing))
+                                            percentMissing = percentMissing,
+                                            ## Variable labels
+                                            varLabels = varLabels))
 
     ## Give a class
     class(TableOneObject) <- "TableOne"

R/ExtractSmd.R

@@ -4,6 +4,8 @@
 ##'
 ##'
 ##' @param x A stratified (svy)TableOne object containing standardized mean differences.
+##' @param varLabels Whether to replace variable names with variable labels obtained from \code{labelled::var_label()} function.
+##'
 ##' @return A vector or matrix containing the average standardized mean differences (if more than two contrasts exist) as well as the all possible pairwise standardized mean differences. Variables are ordered in the same order as the printed table.
 ##' @author Kazuki Yoshida
 ##' @seealso
@@ -13,7 +15,7 @@
 ##' ## See examples for CreateTableOne and svyCreateTableOne
 ##'
 ##' @export
-ExtractSmd <- function(x) {
+ExtractSmd <- function(x, varLabels = FALSE) {
 
     if (class(x)[1] %in% c("TableOne","svyTableOne")) {
 
@@ -22,7 +24,20 @@ ExtractSmd <- function(x) {
                         attr(x$CatTable,  "smd"))
 
         ## Order by table variable order
-        matSmd[x$MetaData$vars,]
+        matSmd <- matSmd[x$MetaData$vars,]
+
+        ## Use variable labels if requested.
+        if (varLabels) {
+            for (i in seq_along(x$MetaData$vars)) {
+                if (!is.null(x$MetaData$varLabels[[i]])) {
+                    ## If the corresponding variable label is non-null replace
+                    rownames(matSmd)[i] <- x$MetaData$varLabels[[i]]
+                }
+            }
+        }
+
+        ## Return manipulated matrix
+        matSmd
 
     } else if (class(x)[1] %in% c("ContTable","svyContTable","CatTable","svyCatTable")) {
 

R/ShowRegTable.R

@@ -46,8 +46,9 @@ ShowRegTable <- function(model, exp = TRUE, digits = 2, pDigits = 3, printToggle
         ## nlme needs special handling
         ## Use column 2 because it is the point estimate
         modelCoef <- nlme::intervals(model)[[1]][, 2]
-    } else if (any(class(model) %in% c("lmerMod","glmerMod"))) {
+    } else if (any(class(model) %in% c("lmerMod","glmerMod","merModLmerTest"))) {
         ## (g)lmer gives confint for other extra parameters
+        ## lmerTest::lmer() gives a merModLmerTest object.
         modelCoef <- coef(summary(model))[,1]
     } else {
         modelCoef <- coef(model)
@@ -58,17 +59,26 @@ ShowRegTable <- function(model, exp = TRUE, digits = 2, pDigits = 3, printToggle
         ## nlme needs special handling
         ## Drop column 2 because it is the point estimate
         modelConfInt <- nlme::intervals(model)[[1]][, -2]
-    } else if (any(class(model) %in% c("lmerMod","glmerMod"))) {
-        ## (g)lmer gives confint for other extra parameters
+    } else if (any(class(model) %in% c("lmerMod","glmerMod","merModLmerTest"))) {
+        ## (g)lmer gives confint for other extra parameters.
+        ## The bottom ones are for fixed effects.
         modelConfInt <- tail(suppressMessages(ciFun(model)), length(modelCoef))
     } else {
         modelConfInt <- suppressMessages(ciFun(model))
     }
 
-    ## P-value extraction
+    ## Extract p-values
     if (any(class(model) %in% c("gls", "lme"))) {
         ## nlme needs special handling
         modelSummaryMat <- summary(model)$tTable
+    } else if (any(class(model) %in% c("lmerMod"))) {
+        ## lmerMod does not have p-values, add NA's
+        modelSummaryMat <- coef(summary(model))
+        modelSummaryMat <- cbind(modelSummaryMat,
+                                 rep(NA, nrow(modelSummaryMat)))
+    } else if (any(class(model) %in% c("merModLmerTest"))) {
+        ## Need to specify explicitly to invoke the correct summary method.
+        modelSummaryMat <- coef(lmerTest::summary(model))
     } else {
         modelSummaryMat <- coef(summary(model))
     }

R/modules-print.R

@@ -322,7 +322,7 @@ ModuleContFormatStrata <- function(ContTable, nVars, listOfFunctions, digits) {
 ################################################################################
 
 ## Module to loop over variables within a stratum formatting categorical variables
-ModuleCatFormatVariables <- function(lstVars, varsToFormat, fmt, level, cramVars, showAllLevels) {
+ModuleCatFormatVariables <- function(lstVars, varsToFormat, fmt, level, cramVars, dropEqual, showAllLevels) {
 
     ## Loop over variables within a stratum
     ## Each list element is a data frame summarizing levels
@@ -378,9 +378,12 @@ ModuleCatFormatVariables <- function(lstVars, varsToFormat, fmt, level, cramVars
                        DF[1,"crammedRowInd"] <- "crammed"
 
                    } else {
-                       ## Otherwise, keep the second level only
-                       ## change variable name, and delete the first level.
-                       DF$var <- sprintf("%s = %s", DF$var, DF$level)
+                       ## Otherwise, keep the second level only.
+                       ## Change variable name if dropEqual = FALSE.
+                       if (!dropEqual) {
+                           DF$var <- sprintf("%s = %s", DF$var, DF$level)
+                       }
+                       ## Delete the first level.
                        DF <- DF[-1, , drop = FALSE]
                    }
 
@@ -412,7 +415,7 @@ ModuleCatFormatVariables <- function(lstVars, varsToFormat, fmt, level, cramVars
 
 
 ## Module to loop over strata formatting categorical variables
-ModuleCatFormatStrata <- function(CatTable, digits, varsToFormat, cramVars, showAllLevels) {
+ModuleCatFormatStrata <- function(CatTable, digits, varsToFormat, cramVars, dropEqual, showAllLevels) {
 
     ## Create format for percent used in the loop
     fmt1 <- paste0("%.", digits, "f")
@@ -433,6 +436,7 @@ ModuleCatFormatStrata <- function(CatTable, digits, varsToFormat, cramVars, show
                                             varsToFormat  = varsToFormat,
                                             fmt           = fmt1,
                                             cramVars      = cramVars,
+                                            dropEqual     = dropEqual,
                                             showAllLevels = showAllLevels)
 
 
@@ -632,7 +636,7 @@ ModuleFormatTables <- function(x, catDigits, contDigits,
                                explain, pDigits,
                                ## print.CatTable arguments passed
                                format, exact,
-                               showAllLevels, cramVars,
+                               showAllLevels, cramVars, dropEqual,
                                ## print.ContTable arguments passed
                                nonnormal, minMax, insertLevel
                                ) {
@@ -680,6 +684,7 @@ ModuleFormatTables <- function(x, catDigits, contDigits,
                      format = format, exact = exact,
                      showAllLevels = showAllLevels,  # Returns one more column if TRUE
                      cramVars = cramVars,
+                     dropEqual = dropEqual,
 
                      ## print.ContTable arguments passed
                      nonnormal = nonnormal, minMax = minMax, insertLevel = showAllLevels

R/print.CatTable.R

@@ -13,6 +13,7 @@
 ##' @param format The default is "fp" frequency (percentage). You can also choose from "f" frequency only, "p" percentage only, and "pf" percentage (frequency).
 ##' @param showAllLevels Whether to show all levels. FALSE by default, i.e., for 2-level categorical variables, only the higher level is shown to avoid redundant information.
 ##' @param cramVars A character vector to specify the two-level categorical variables, for which both levels should be shown in one row.
+##' @param dropEqual Whether to drop " = second level name" description indicating which level is shown for two-level categorical variables.
 ##' @param test Whether to show p-values. TRUE by default. If FALSE, only the numerical summaries are shown.
 ##' @param exact A character vector to specify the variables for which the p-values should be those of exact tests. By default all p-values are from large sample approximation tests (chisq.test).
 ##' @param smd Whether to show standardized mean differences. FALSE by default. If there are more than one contrasts, the average of all possible standardized mean differences is shown. For individual contrasts, use \code{summary}.
@@ -95,6 +96,7 @@ function(x,                        # CatTable object
          format        = c("fp","f","p","pf")[1], # Format f_requency and/or p_ercent
          showAllLevels = FALSE,
          cramVars      = NULL,     # variables to be crammed into one row
+         dropEqual     = FALSE,    # Do not show " = second level" for two-level variables
 
          test          = TRUE,     # Whether to add p-values
          exact         = NULL,     # Which variables should be tested with exact tests
@@ -163,6 +165,7 @@ function(x,                        # CatTable object
                           digits        = digits,
                           varsToFormat  = varsToFormat,
                           cramVars      = cramVars,
+                          dropEqual     = dropEqual,
                           showAllLevels = showAllLevels)
 
 

R/print.TableOne.R

@@ -14,9 +14,11 @@
 ##' @param smd Whether to show standardized mean differences. FALSE by default. If there are more than one contrasts, the average of all possible standardized mean differences is shown. For individual contrasts, use \code{summary}.
 ##' @param noSpaces Whether to remove spaces added for alignment. Use this option if you prefer to align numbers yourself in other software.
 ##' @param padColnames Whether to pad column names with spaces to center justify. The default is FALSE. It is not conducted if noSpaces = TRUE.
+##' @param varLabels Whether to replace variable names with variable labels obtained from \code{labelled::var_label()} function.
 ##' @param format The default is "fp" frequency (percentage). You can also choose from "f" frequency only, "p" percentage only, and "pf" percentage (frequency).
 ##' @param showAllLevels Whether to show all levels. FALSE by default, i.e., for 2-level categorical variables, only the higher level is shown to avoid redundant information.
 ##' @param cramVars A character vector to specify the two-level categorical variables, for which both levels should be shown in one row.
+##' @param dropEqual Whether to drop " = second level name" description indicating which level is shown for two-level categorical variables.
 ##' @param exact A character vector to specify the variables for which the p-values should be those of exact tests. By default all p-values are from large sample approximation tests (chisq.test).
 ##' @param nonnormal A character vector to specify the variables for which the p-values should be those of nonparametric tests. By default all p-values are from normal assumption-based tests (oneway.test).
 ##' @param minMax Whether to use [min,max] instead of [p25,p75] for nonnormal variables. The default is FALSE.
@@ -43,11 +45,13 @@ function(x,                   # TableOne object
          smd           = FALSE, # Whether to add standardized mean differences
          noSpaces      = FALSE, # Whether to remove spaces for alignments
          padColnames   = FALSE, # Whether to pad column names for alignments
+         varLabels     = FALSE, # Whether to show variable labels instead of names.
 
          ## Categorical options
          format        = c("fp","f","p","pf")[1], # Format f_requency and/or p_ercent
          showAllLevels = FALSE, # Show all levels of a categorical variable
          cramVars      = NULL,  # Which 2-level variables to show both levels in one row
+         dropEqual     = FALSE, # Do not show " = second level" for two-level variables
          exact         = NULL,  # Which variables should be tested with exact tests
 
          ## Continuous options
@@ -67,6 +71,7 @@ function(x,                   # TableOne object
                                     ## Returns one more column if TRUE
                                     showAllLevels = showAllLevels,
                                     cramVars = cramVars,
+                                    dropEqual = dropEqual,
 
                                     ## print.ContTable arguments passed
                                     nonnormal = nonnormal, minMax = minMax,
@@ -91,10 +96,32 @@ function(x,                   # TableOne object
     spcFmtEltTables <- ModuleAddSpacesToTable(FmtElementTables, nSpacesToAdd, showAllLevels)
 
 
-    ## Create a list of one variable tables excluding sample size row
+    ## Create a list of one variable tables excluding sample size row.
+    ## This is based on the variable order in the MetaData.
     lstOneVarTables <- ModuleListOfOneVarTables(spcFmtEltTables,
                                                 MetaData = x$MetaData)
 
+    ## Replace variable names with variable labels if requested.
+    ## Loop over the variable replacing its name with its label.
+    if (varLabels) {
+        lstOneVarTables <-
+            lapply(seq_along(lstOneVarTables),
+                   function(i) {
+                       ## Each element is a string matrix.
+                       mat <- lstOneVarTables[[i]]
+                       ## Manipulate if a non-NULL label is available.
+                       ## Note MetaData$varLabels is a list.
+                       if (!is.null(x$MetaData$varLabels[[i]])) {
+                           ## The first row name contains the variable name
+                           ## without preceding space. Replace by exact matching.
+                           rownames(mat)[1] <- gsub(paste0("^", x$MetaData$vars[i]),
+                                                    x$MetaData$varLabels[[i]],
+                                                    rownames(mat)[1])
+                       }
+                       ## Return the entire matrix.
+                       mat
+                   })
+    }
 
     ## Check if the first row is CatTable element
     ## if so, pick sample size row from space-padded CatTable element

R/print.svyCatTable.R

@@ -13,6 +13,7 @@
 ##' @param format The default is "fp" frequency (percentage). You can also choose from "f" frequency only, "p" percentage only, and "pf" percentage (frequency).
 ##' @param showAllLevels Whether to show all levels. FALSE by default, i.e., for 2-level categorical variables, only the higher level is shown to avoid redundant information.
 ##' @param cramVars A character vector to specify the two-level categorical variables, for which both levels should be shown in one row.
+##' @param dropEqual Whether to drop " = second level name" description indicating which level is shown for two-level categorical variables.
 ##' @param test Whether to show p-values. TRUE by default. If FALSE, only the numerical summaries are shown.
 ##' @param exact This option is not available for tables from weighted data.
 ##' @param smd Whether to show standardized mean differences. FALSE by default. If there are more than one contrasts, the average of all possible standardized mean differences is shown. For individual contrasts, use \code{summary}.
@@ -40,6 +41,7 @@ function(x,                        # CatTable object
          format        = c("fp","f","p","pf")[1], # Format f_requency and/or p_ercent
          showAllLevels = FALSE,
          cramVars      = NULL,     # variables to be crammed into one row
+         dropEqual     = FALSE,    # Do not show " = second level" for two-level variables
 
          test          = TRUE,     # Whether to add p-values
          exact         = NULL,     # Which variables should be tested with exact tests
@@ -109,6 +111,7 @@ function(x,                        # CatTable object
                           digits        = digits,
                           varsToFormat  = varsToFormat,
                           cramVars      = cramVars,
+                          dropEqual     = dropEqual,
                           showAllLevels = showAllLevels)