Creating a running patient id number

[Girish Venkataraman]

I need created a patientid column for each unique accession number. I sense something with foreach or forvalue with some incrementing. I tried _n but that only gives incremental numbers to duplicate accessions with bysort accession. There are some patients with duplicate biopsies that need have the same patientid.
This sounds pretty simple but I am struggling with it.

************************************************** *********************

* Example generated by -dataex-. For more info, type help dataex
clear
input str9 accession str5 casesource

 "1"         "Japan"
"10"        "Japan"
"10750"     "UIC"  
"145"       "UIC"  
"146"       "UIC"  
"148"       "UIC"  
"148"       "UIC"  
"152"       "UIC"  
"153"       "UIC"  
"154"       "UIC"  
"156"       "UIC"  
"157"       "UIC"  
"2"         "Japan"
"3"         "Japan"
"4"         "Japan"
"42195"     "UIC"  
"4"         "Japan"
"6"         "Japan"
"7"         "Japan"
"8"         "Japan"
"9"         "Japan"
"D21-4179"  "UOC"  
"M"         "UOC"  
"NONE"      "UOC"  
"PO21-7775" "UOC"  
"PO21-8645" "UOC"  
"S10-18829" "UOC"  
"S10-236"   "UOC"  
"S10-5305"  "UOC"  
"S11-14272" "UOC"  
"S11-15357" "UOC"  
"S11-21217" "UOC"  
"S11-28268" "UOC"  
"S14-1306"  "UOC"  
"S14-18381" "UOC"  
"S15-10912" "UOC"  
"S16-10722" "UOC"  
"S16-29371" "UOC"  
"S17-10741" "UOC"  
"S17-15971" "UOC"  
"S17-9"     "UOC"  
"S19-27538" "UOC"  
"S19-27598" "UOC"  
"S19-28431" "UOC"  
"S19-28814" "UOC"  
"S19-3466"  "UOC"  
"S20-13935" "UOC"  
"S20-14825" "UOC"  
"S20-15844" "UOC"  
"S20-17289" "UOC"  
"S20-18545" "UOC"  
"S20-19738" "UOC"  
"S20-27266" "UOC"  
"S20-31227" "UOC"  
"S20-379"   "UOC"  
"S21-10855" "UOC"  
"S21-17350" "UOC"  
"S21-19659" "UOC"  
"S21-208"   "UOC"  
"S21-25162" "UOC"  
"S21-25729" "UOC"  
"S21-27767" "UOC"  
"S21-29834" "UOC"  
"S21-30841" "UOC"  
"S21-31882" "UOC"  
"Unknown"   "UOC"  
"s15-20074" "UOC"  
"s18-6176"  "UOC"  
"s20-11917" "UOC"  
"s20-18323" "UOC"  
"s20-18534" "UOC"  
"s20-22472" "UOC"  
"s20-2674"  "UOC"  
"s20-29073" "UOC"  
"s20-32003" "UOC"  
"s20-4895"  "UOC"  
"s21-659"   "UOC"  
end

[Jared Greathouse]

egen id = group(accession)

[Girish Venkataraman]

That was so simple. Will read more on egen. Thanks Jared.

[Nick Cox]

With many, many patients make sure that you go

 
 egen long id = group(accession)

With not so many patients, you could keep the original values as value labels with

  
 egen id = group(accession), label 

[Girish Venkataraman]

That is very useful, Nick. I think that would be appropriate for me too. Usually final datasets are not more than 200 patients.