storing -egen varlist- after each regression

[Stephen Ch]

Consider the following panel dataset:

id	datadate	affiliation	revenue	workforce
100	31-May-01	NBA	200	11
100	31-May-02	NBA	250	15
101	31-May-03	NHL	900	20
101	31-May-01	NFL	932	9
102	31-May-02	NHL	400	65
103	31-May-03	FIFA	230	25
103	31-May-04	FIFA	890	28

I am running the following regression:

reg revenue workforce

After this regression, I would like to count distinct affiliation and ids in and out of my regression sample, and this is my code:

egen distinct_id = group(id) if e(sample)
egen distinct_affiliation = group(affiliation) if e(sample)
sum distinct_id distinct_affiliation

My understanding is that the "max" of the -sum- gives you the distinct counts for each variable of my interest.

I am running several regressions (approx. 30), but after each regression, I would like to store the "max" number from the sum and export to Excel as csv or xlsx file.

Do you know what is the best way to go about doing this?

[Demetrio Guzzardi]

Using

gen max=`r(max)'

after the sum command, you save the maximum value in a variable.

[Stephen Ch]

Great, thanks.

How would I store -max- in some sort of a matrix and export to csv?

Using

gen max=`r(max)'

after the sum command, you save the maximum value in a variable.

[Stephen Ch]

Hmmm, the command does not work to store the information...any other suggestions?
How can I store this information and export to Excel?

Using

gen max=`r(max)'

after the sum command, you save the maximum value in a variable.

[Andrew Musau]

You show only 1 regression and there is no scope to loop your code. Spend more time creating a reproducible example with at least 2 regressions and specify exactly what statistics you want. Also, see FAQ Advice #12 on how to use the dataex command to present data examples.

[Stephen Ch]

Hi Andrew,

This is helpful. I will use -dataex- and put a proper formatted data when asking a question.

Thanks!

You show only 1 regression and there is no scope to loop your code. Spend more time creating a reproducible example with at least 2 regressions and specify exactly what statistics you want. Also, see FAQ Advice #12 on how to use the dataex command to present data examples.