Re: st: replacing with mean

[Fabio Zona <fabio.zona@unibocconi.it>]

Dimitriy!
Thank you soo much!! It works perfectly now!

I thought it would be a general command for all variables; now I understand that I need to run the command for every variable forwhich I want to calculate the median, but this is perfectly manageable! So your command is correct and works perfectly!

Again, thank you very very much!

Fabio





----- Messaggio originale -----
Da: "Dimitriy V. Masterov" <dvmaster@gmail.com>
A: statalist@hsphsun2.harvard.edu
Inviato: Giovedì, 2 dicembre 2010 21:25:51 GMT +01:00 Amsterdam/Berlino/Berna/Roma/Stoccolma/Vienna
Oggetto: Re: st: replacing with mean

Fabio,

Since you have your own data, you should use:

levelsof listA, local(listA)

foreach var of varlist Var* {
       qui gen `var'_med=.

       foreach firm in `listA' {
               qui gen inc=cond(listA != `firm',1,.)
               qui egen med=median(`var'*inc), by(industry)
               qui replace `var'_med=med if listA==`firm'
               qui drop med inc
       }

}

This code assumes that you have data that looks like this:

industry listA	Var1	Var2	Var3
A	1	0	1	1
A	2	0	1	1
A	3	1	1	1
A	4	1	0	0
A	5	1	0	0
A	6	1	0	2
A	7	1	1	2
A	8	0	0	2
A	9	1	1	1

I was trying to use the variable names from your original e-mail as
much as I could. If your variables have different names (including
capitalization: Var1 is not the same as var1), you many need to modify
my code to reflect that.

DVM

On Thu, Dec 2, 2010 at 3:12 PM, Fabio Zona <fabio.zona@unibocconi.it> wrote:
> Hi Dimitry, that's great! Thank you very much!
>
> I tried to implement your command; however, I get an error as it says:
>
>
> variable Var* not found
> r(111);
>
>
> Might it depend on the fact that I start your procedure from here:
>
>  levelsof listA, local(listA)
>
> as I already have my dataset ready for calculation)?
>
> F
>
>
>
>
> ----- Messaggio originale -----
> Da: "Dimitriy V. Masterov" <dvmaster@gmail.com>
> A: statalist@hsphsun2.harvard.edu
> Inviato: Giovedì, 2 dicembre 2010 20:41:50 GMT +01:00 Amsterdam/Berlino/Berna/Roma/Stoccolma/Vienna
> Oggetto: Re: st: replacing with mean
>
> Fabio,
>
> Nick's suggestion can be implemented in the 2nd loop below:
>
> ************************************************************
> clear all
> set more off
>
> /* Fake Data */
> set obs 26
>
> gen str1 industry=""
>
> local c=1
> foreach l in `c(ALPHA)' {
>        replace industry = "`l'" in `c'
>        local ++c
> }
>
> expand 27
>
> sort industry
>
> forvalues v=1/20 {
>        gen Var`v'=round(uniform()*`v',1)
> }
>
> gen listA=_n
>
>
> /* Actual Loop */
> levelsof listA, local(listA)
>
> foreach var of varlist Var* {
>        qui gen `var'_med=.
>
>        foreach firm in `listA' {
>                qui gen inc=cond(listA != `firm',1,.)
>                qui egen med=median(`var'*inc), by(industry)
>                qui replace `var'_med=med if listA==`firm'
>                qui drop med inc
>        }
>
> }
> **************************************************************
> *
> *   For searches and help try:
> *   http://www.stata.com/help.cgi?search
> *   http://www.stata.com/support/statalist/faq
> *   http://www.ats.ucla.edu/stat/stata/
>
> *
> *   For searches and help try:
> *   http://www.stata.com/help.cgi?search
> *   http://www.stata.com/support/statalist/faq
> *   http://www.ats.ucla.edu/stat/stata/
>

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

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