Re: st: RE: compare files via merge

[David Kantor <kantor.d@att.net>]

Hello Rasool,
Thanks very much. I will look into your suggestions.
--David

At 06:09 AM 1/3/2014, you wrote:
> Dear David,
>
> I have made an ado file for generating list if id discrepancies in 
> two data files. May be this will help you. Copy the following code 
> and save an ado file in your working directory. Then run as:
>
> ----cmpids file1.dta file2.dta idvar----
>
> You can change the program's drop statement if you want the list of 
> matched ones.
> Change this line
>         qui drop if count_one == 1 & count_two == 1
> with this one.
>         qui drop if count_one ~= 1 & count_two ~= 1
>
> Best of luck.
> Rasool Bux
>
>
> *********program begin ************
> capture program drop cmpids
> program define cmpids
>         args fone ftwo idv
>
>                 if `"`fone'"' == "" | `"`ftwo'"' == "" {
>                 di as err "Two data files are required; nothing to do"
>                 exit 498
>         }
>
>                 if "`fone'" == "`ftwo'" {
>                         di as err "Same file is not used for 
> structure compare"
>                         exit 498
>                 }
>         else {
>                                 local idvar "`idv'"
>                                 if substr("`fone'", -4, 4) != 
> ".dta" local fone "`fone'.dta"
>                                 if substr("`ftwo'", -4, 4) != 
> ".dta" local ftwo "`ftwo'.dta"
>
>                 qui use "`fone'", clear
>                                 qui contract `idvar', freq(count_one) nomiss
>                                 qui label variable count_one "Freq in file 1"
>                                 qui save freq1.dta, replace
>
>                 qui use "`ftwo'", clear
>                                 qui contract `idvar', freq(count_two) nomiss
>                                 qui label variable count_two "Freq in file 2"
>                                 qui save freq2.dta, replace
>
>                                 qui use freq1.dta,clear
>                                 qui merge 1:1 `idvar' using freq2.dta, norep
>                                 qui replace count_one = 0 if count_one==.
>                                 qui replace count_two = 0 if count_two==.
>                                 qui drop if count_one == 1 & count_two == 1
>                                 qui drop _merge
>                                 qui save freqtbl.dta, replace
>                                 qui erase freq1.dta
>                                 qui erase freq2.dta
>                                 brow `idvar' count_one count_two
>                         }
> end
>
> ****************** program end **********************
> [...]

*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/faqs/resources/statalist-faq/
*   http://www.ats.ucla.edu/stat/stata/