Running regressions in loop and storing results

[Anton Ivanov]

Hello!

The example of my data is provided below. My goal is to run a series of simple regressions for each of the individual asin codes (see examples below) and for each of the runs store the beta coefficient, standard error, and p-value in separate columns. What would be the appropriate way to do so?

Thanks!

reg overall month if asin == "373250460"
reg overall month if asin == "446697192"
///etc.

* Example generated by -dataex-. For more info, type help dataex
clear
input str10 asin int month double overall
"373250460"  528         4
"373250460"  530         5
"373250460"  531         5
"373250460"  534         3
"373250460"  535         1
"373250460"  536         2
"373250460"  540         3
"373250460"  541         4
"373250460"  543         4
"373250460"  544         1
"373250460"  545         1
"373250460"  547         1
"373250460"  548         1
"373250460"  551         4
"373250460"  553       1.5
"373250460"  555         1
"373250460"  564         3
"373250460"  567         4
"373250460"  568         4
"373250460"  569         2
"373250460"  579         2
"373250460"  587         3
"373250460"  591         4
"373250460"  629         3
"373250460"  634       4.5
"373250460"  677         4
"446697192"  594 3.9000001
"446697192"  595 4.3333335
"446697192"  596         3
"446697192"  598         5
"446697192"  601         4
"446697192"  605         3
"446697192"  611         3
"446697192"  614         4
"446697192"  624         2
"446697192"  626         4
"446697192"  628         4
"446697192"  672         4
"575074205"  567 3.6666667
"575074205"  574         4
"575074205"  583         4
"575074205"  656         4
"575074205"  670         2
"575074205"  687         3
"812566564"  478         5
"812566564"  479         4
"812566564"  483         4
"812566564"  494         1
"812566564"  504         4
"812566564"  537         2
"812566564"  630         2
"812566564"  631         4
"812566564"  650         5
"812566564"  653         5
"812566564"  666         5
"812566564"  681         5
"979526191"  610         4
"979526191"  611         5
"979526191"  612      3.25
"979526191"  615         5
"979526191"  616         5
"979526191"  619         4
"979526191"  633         1
"979526191"  642         1
"979526191"  643         1
"979526191"  645         3
"1039869017" 638         5
"1039869017" 644         5
"1059983354" 654         2
"1059983354" 657         1
"1059983354" 667         4
"1059983354" 669         5
"1060078031" 643 4.8000002
"1060078031" 644         4
"1060078031" 645 4.0833335
"1060078031" 646 2.8666666
"1060078031" 647         4
"1060078031" 648 4.1333332
"1060078031" 649 4.5384617
"1060078031" 650 4.2142859
"1060078031" 651 3.3636363
"1060078031" 652 4.2222223
"1060078031" 653 3.3333333
"1060078031" 654 3.8888888
"1060078031" 655       2.5
"1060078031" 656 3.4000001
"1060078031" 657         3
"1060078031" 658         5
"1060078031" 661         3
"1060078031" 677         5
"1060240718" 659         1
"1060240718" 678         1
"1061790827" 659         5
"1061790827" 660         5
"1061790827" 662         3
"1061790827" 663         1
"1061790827" 665         5
"1400532620" 606         5
"1400532620" 607         3
"1400532620" 608 4.8333335
end

[Clyde Schechter]

The simplest way to do this is:

rangestat (reg) overall month, by(asin) interval(overall . .)

-rangestat- is by Robert Picard, Nick Cox, and Roberto Ferrer. It is available from SSC.

[Anton Ivanov]

Thank you Clyde Schechter

[Attaullah Shah]

ssc install asreg

. bys asin: asreg overall month

. list in 1/10

     +----------------------------------------------------------------------------------------+
     |       asin   month     overall   _Nobs         _R2      _adjR2    _b_month     _b_cons |
     |----------------------------------------------------------------------------------------|
  1. | 1039869017     644           5     100   .02311415   .01314592   .00381027   1.1089697 |
  2. | 1039869017     638           5     100   .02311415   .01314592   .00381027   1.1089697 |
  3. | 1059983354     667           4     100   .02311415   .01314592   .00381027   1.1089697 |
  4. | 1059983354     654           2     100   .02311415   .01314592   .00381027   1.1089697 |
  5. | 1059983354     657           1     100   .02311415   .01314592   .00381027   1.1089697 |
     |----------------------------------------------------------------------------------------|
  6. | 1059983354     669           5     100   .02311415   .01314592   .00381027   1.1089697 |
  7. | 1060078031     643   4.8000002     100   .02311415   .01314592   .00381027   1.1089697 |
  8. | 1060078031     652   4.2222223     100   .02311415   .01314592   .00381027   1.1089697 |
  9. | 1060078031     646   2.8666666     100   .02311415   .01314592   .00381027   1.1089697 |
 10. | 1060078031     657           3     100   .02311415   .01314592   .00381027   1.1089697 |
     +----------------------------------------------------------------------------------------+

More on asreg can be found here https://fintechprofessor.com/2017/12...ions-in-stata/

[Anton Ivanov]

Thank you Attaullah Shah This is a great suggestion.