Assignee browyn hall

TestOnDisambig.md

@@ -0,0 +1,7 @@
+## TEST ON REMOTE ENV
+
+1. ssh PVDis
+2. `conda activate disambig`
+3. `export PYTHONPATH="/home/centos/ba_disambiguation_20230330:$PYTHONPATH"`
+4. `export DISAMBIGUATION_ROOT=/home/centos/ba_disambiguation_20230330`
+5. ``python assignee_clustering_runner.py``

augmenting_data/namestandarization programs20230717123820 (1)/combabbrev.do

@@ -0,0 +1,35 @@
+* combabbrev.do
+
+* 1/2007 JBessen
+*
+* combine single char sequences in standard_name
+* this assumes name string begins and ends with space
+
+gen str90 outname = " "
+
+* remove quote characters
+replace standard_name = subinstr( standard_name, `"""',  "", 30)
+
+local i = 1
+while `i' <= _N {
+
+	* do next name
+	local j = 1
+	local len = wordcount(standard_name[`i']) 
+
+	* do next word	
+	while `j' <= `len' {
+
+		local outword = word( standard_name[`i'], `j')
+		if  strlen("`outword'")~=1 | strlen(word( standard_name[`i'],`j'+1))!=1  {
+			local outword = "`outword'"+" "
+		}
+		qui replace outname = outname + "`outword'" in `i'
+
+		local j = `j' + 1
+	}
+	local i = `i' + 1
+}
+
+replace standard_name = outname
+drop outname

augmenting_data/namestandarization programs20230717123820 (1)/corpentities.do

@@ -0,0 +1,164 @@
+* corpentities.do
+* 
+* J Bessen 1/2007
+* makes list of unique corporate entities
+* unique by standard_name, (not stem_name)
+*
+*	takes 2005 compustat header (cshdr05.dta)
+*		generates standard_names and stem_names
+*		flags duplicate standard_names for elimination
+*		matches standard_names against standard_names to find similar names
+*		allows manual assignment of similar names as precursors/subsidiaries (selfassign.dta)
+* output
+*	- cstatnames.dta
+*		stat = status code
+*			= 0, eliminate
+*			= 1, precursor/subsidiary
+*			= 2,. successor/parent
+*	- cstat.txt, unique list of standard_names for matching
+
+*------------------
+* site specific code
+* -----------------
+clear
+global CSDIR ="C:/docume~1/HP_Owner/mydocu~1/data/pdp/freqmatch"
+global NAMDIR ="C:/docume~1/HP_Owner/mydocu~1/data/pdp/freqmatch"
+cap cd $CSDIR
+set memory 150m
+*-------------------
+
+
+* clean company names -------------------------------------------------
+
+insheet using compustat_country_codes.csv,comma clear
+rename code ctrycode
+drop country
+rename uspto country
+label var country "Headquarters country from Compustat"
+sort ctrycode
+save temp,replace
+
+use cshdr05,clear
+rename finc ctrycode
+sort ctrycode
+merge ctrycode using temp,nokeep
+tab _m
+drop _m
+replace country = "US" if ctrycode==. | ctrycode==0
+drop ctrycode
+tab country
+
+do nameonly_main
+save cstatnames,replace
+
+* eliminate duplicate standard names ----------------------------------
+* 
+* COMPARISON RULES
+* if firstyr>= and lastyr< => eliminate
+* if firstyr= and lastyr = => take one with >sales, eliminate other
+* if firstyr< and lastyr<  => make precursor
+* NOTE: the sales rule will eliminate pre-FASB 94 records
+
+sort standard_name
+by standard_name: egen dup=count(firstyr)
+keep if dup>1
+gsort standard_name lastyr firstyr -sales
+by standard_name: egen maxyr = max(lastyr)
+gen mintmp = firstyr if lastyr==maxyr
+by standard_name: egen minyr = min(mintmp) 
+drop mintmp
+gen double mintmp = sales if lastyr==maxyr & minyr==firstyr
+by standard_name: egen maxsls=max(mintmp)
+
+gen stat = 0 /* eliminate */
+replace stat=1 if firstyr < minyr	/* precursor */
+replace stat=2 if firstyr==minyr & lastyr==maxyr & maxsls<.01+sales
+
+
+keep standard_name gvkey stat firstyr lastyr
+sort gvkey
+save dupnames,replace
+
+* now do self matches---------------------------------------------------
+* first, output list of names for freqmatch2.pl
+
+use cstatnames,clear
+sort gvkey
+merge gvkey using dupnames
+drop if stat==0 | stat==1
+outsheet stem_name standard_name gvkey using selfmatch.txt,replace noq noname
+
+* RUN MANUAL MAtch
+*
+* perl freqmatch2.pl selfmatch.txt selfmatch.txt >self.csv
+*
+* onetime update using "update self.do"
+* creates selfassign.dta
+* 
+
+log using cstatnames.log,text replace
+use dupnames,clear
+* special code
+replace stat=0 if gvkey==139901 /* eliminate this one for UBS */
+
+sort stand stat
+gen gvtmp = gvkey if stat==2
+by stand: egen pdpco= max(gvtmp)
+
+keep gvkey pdpco stat
+sort gvkey
+
+merge gvkey using selfassign,replace update
+tab _merge
+replace stat=1 if stat==.
+sort gvkey
+gen dup=gvkey==gvkey[_n-1]
+tab dup
+keep gvkey pdpco stat
+save temp,replace
+
+use cstatnames,clear
+sort gvkey
+merge gvkey using temp
+tab _merge
+tab stat
+
+* special code...now in corpfix.csv
+*replace stat = 1 if gvkey==17346 | gvkey==101410 /* ABB */
+*replace pdpco = 210418 if gvkey==17346 | gvkey==101410 /* ABB */
+*replace stat = 1 if gvkey==117298 | gvkey==118653 | gvkey==121742 | gvkey==143176 /* genzyme */
+*replace pdpco=12233 if gvkey==117298 | gvkey==118653 | gvkey==121742 | gvkey==143176 /* genzyme */
+*replace stat=1 if gvkey==66013 /* georgia pacific */
+*replace pdpco=5134 if gvkey==66013 /* georgia pacific */
+*replace stat=1 if gvkey==124015 /* quantum */
+*replace pdpco=8867 if gvkey==124015 /* quantum */
+
+
+***drop if stat==0
+replace pdpco = gvkey if pdpco==.
+replace stat=2 if stat==.
+label variable stat "0=ignore, 1=precursor/subsidiary, 2=successor/parent"
+label variable pdpco "PDP corporate entity key"
+sort pdpco
+drop _merge
+save cstatnames,replace
+
+* manual fixes
+clear
+insheet using corpfix.csv, comma
+drop comment
+sort gvkey
+save temp,replace
+
+use cstatnames, clear
+sort gvkey
+merge gvkey using temp, replace update
+tab _m
+drop _merge
+save cstatnames,replace
+
+sort standard_name
+keep if gvkey==pdpco
+outsheet stem_name standard_name pdpco using cstat.txt,replace noq noname
+log close
+