loop over the panel IDs to use quantile regressions xtqreg

[Aamina Khurraa]

Hi,

I am running quantile regression for panel of 260 firms for period of 15 years. The general regression code xtqreg y x, gives an aggregate estimation of the entire data with all the panels. However, I want to run this regression for the individual panel IDs and don't find any way around. Joao Silvia suggested to ask Stata experts, how can I do a loop over the panel IDs to get the estimates of the regression for the individual panel IDs?


Best Regards

[Clyde Schechter]

Assuming you want to save the coefficient of x and its standard error as results in your data set, the simplest approach is this:

capture program drop one_id
program define one_id
    qreg y x
    gen coeff = _b[x]
    gen se = _se[x]
    exit
end

runby one_id, by(ID) status

Notes:

1. -runby- is written by Robert Picard and me; it is available from SSC.

2. Note that in program one_id I use qreg, not -xtqreg-. This is because -runby- will process this data one panel at a time. So when one_id is running, there is no longer any panel structure: only one panel is available.

3. I am assuming the the variable ID identifies panels, and that these panels are the same panels you originally used when you -xtset- your data before running -xtqreg-. If this is not true, post back with an explanation and also use the -dataex- command to show example data.

[Aamina Khurraa]

Dear Clyde,

Thanks for your support. I am pasting the output that I am getting after running the code above:

elapsed ----------- by-groups ---------- ------- observations ------ time
time count errors no-data processed saved remaining
------------------------------------------------------------------------------------
00:00:01 25 25 0 3,907 0 00:00:09
00:00:02 69 69 0 10,365 0 00:00:05
00:00:03 110 110 0 16,321 0 00:00:04
00:00:04 152 152 0 21,571 0 00:00:03
(now reporting every 5 seconds)
00:00:07 262 262 0 38,101 0 00:00:00

--------------------------------------
Number of by-groups = 262
by-groups with errors = 262
by-groups with no data = 0
Observations processed = 38,101
Observations saved = 0
--------------------------------------


And when I browse, all the data and variables are gone, its a blank doc with no data. what does this mean? Moreover, I am not sure how meaningful are the results in the output window. they are just the count of observations and errors. I wanted to have the coefficients of all the individual panels through running quantile regression, which I dont's see and find out how to extract them.

Best Regards

[Clyde Schechter]

In order to troubleshoot this, I need a sample of your data. (The code runs with no errors and produces correct results on the grunfeld.dta set provided by StataCorp with suitable renaming of variables, so there is something different about your data.) Use the -dataex- command to show an excerpt from your Stata data set. Be sure that the excerpt you show includes several different IDs and all of the variables that you use in your -xtqreg- (or, with my code, -qreg-) command.

If you are running version 16 or a fully updated version 15.1 or 14.2, -dataex- is already part of your official Stata installation. If not, run -ssc install dataex- to get it. Either way, run -help dataex- to read the simple instructions for using it. -dataex- will save you time; it is easier and quicker than typing out tables. It includes complete information about aspects of the data that are often critical to answering your question but cannot be seen from tabular displays or screenshots. It also makes it possible for those who want to help you to create a faithful representation of your example to try out their code, which in turn makes it more likely that their answer will actually work in your data.

[Aamina Khurraa]

Hi Clyde,

I tried but I am not able to used Dataex with this large T data for each panel ID. I tried to generate random data out of the sample using randomtag, but it would give an error syntax saying type mismatch after the first code, to generate pick

. randomtag if length(y_var) == 4, count(10) gen(pick)

. dataex y_var x_var if pick

thus I am not able to depict all the observations for atleast 2 or 3 panels using dataex, that's why I copied the data as below:

Time	FI_n	Xi(y)	L_MTbills	L_MktR	L_EV
2011m3	2	1399.921	0	-5.670	4.609
2011m4	2	1469.539	0	-2.904	4.015
2011m5	2	1469.539	0.700	3.528	3.587
2011m6	2	1472.498	0.840	-2.225	3.725
2011m7	2	1538.159	0	-2.645	3.623
2011m8	2	1538.159	1	-2.959	3.536
2011m9	2	1541.468	0.930	-4.049	3.558
2011m10	2	1615.177	0	0.719	3.441
2011m11	2	1615.177	0.920	1.472	3.294
2011m12	2	1615.177	0	-1.841	3.307
2012m1	2	2053.130	0	0.045	3.306
2012m2	2	2053.130	0.720	0.940	3.054
2012m3	2	2053.130	0	4.299	3.007
2012m4	2	2217.788	0	0.620	3.184
2012m5	2	2217.788	0.760	3.249	2.677
2012m6	2	2217.788	.	-2.783	2.746
2012m7	2	792.354	.	-6.756	2.770
2012m8	2	792.354	.	-1.196	3.082
2012m9	2	792.354	.	2.455	2.874
2012m10	2	787.717	.	2.028	2.932
2012m11	2	355.369	.	-3.673	2.970
2012m12	2	654.850	.	3.023	3.015
2013m1	2	865.750	.	-0.163	3.123
2013m2	2	1093.095	.	5.105	2.915
2013m3	2	1073.190	.	3.437	3.083
2013m4	2	1103.127	.	3.915	3.078
2013m5	2	919.176	.	10.028	3.148
2013m6	2	1017.167	.	11.072	3.705
2013m7	2	1083.820	.	-6.568	4.199
2013m8	2	906.396	.	3.754	4.387
2013m9	2	936.887	.	-5.575	4.414
2013m10	2	1109.634	.	1.746	4.663
2013m11	2	929.986	.	2.284	4.616
2013m12	2	969.215	.	-2.046	4.611
2014m1	2	805.993	.	-3.078	4.669
2014m2	2	780.052	.	2.696	4.707
2014m3	2	761.091	.	-0.816	4.721
2014m4	2	752.463	.	-1.571	4.710
2014m5	2	655.990	.	-2.205	4.673
2014m6	2	589.394	.	-1.586	4.410
2014m7	2	673.029	.	-4.483	4.421
2014m8	2	729.991	.	2.261	4.565
2014m9	2	720.554	.	4.116	4.568
2014m10	2	665.440	.	2.538	4.507
2014m11	2	598.092	.	-3.470	4.498
2014m12	2	534.727	.	-8.631	4.598
2015m1	2	567.922	.	-3.268	4.931
2015m2	2	558.627	.	0.558	4.939
2015m3	2	491.337	.	0.443	4.865
2015m4	2	516.265	.	-4.952	4.313
2015m5	2	474.058	.	1.494	3.543
2015m6	2	465.644	.	-1.335	3.382
2015m7	2	449.315	.	-1.433	3.218
2015m8	2	389.505	.	0.815	3.063
2015m9	2	362.365	.	-7.178	3.031
2015m10	2	352.028	.	-1.639	3.211
2015m11	2	387.508	.	0.859	3.208
2015m12	2	351.798	.	0.466	3.220
2016m1	2	322.062	.	-3.280	3.124
2016m2	2	321.063	.	-9.338	3.149
2016m3	2	300.825	.	1.800	3.573
2016m4	2	302.035	.	0.409	3.651
2016m5	2	282.030	.	3.071	3.677
2016m6	2	243.787	.	0.158	3.738
2016m7	2	250.284	.	-0.664	3.672
2016m8	2	231.223	.	1.598	3.475
2016m9	2	227.958	.	-0.576	3.429
2016m10	2	216.230	.	-0.394	3.406
2016m11	2	199.837	.	0.050	3.001
2016m12	2	240.195	.	2.800	2.966
2017m1	2	392.783	.	3.427	3.054
2017m2	2	348.468	.	17.278	3.174
2017m3	2	330.129	.	-0.722	4.836
2017m4	2	348.850	.	3.567	4.835
2017m5	2	308.723	.	-2.688	4.870
2017m6	2	300.169	.	-0.846	4.901
2017m7	2	330.785	.	-0.333	4.907
2017m8	2	325.953	.	1.305	4.614
2017m9	2	321.889	.	0.589	4.586
2017m10	2	293.353	.	-3.130	4.587
2017m11	2	242.059	.	-2.515	4.663
2017m12	2	235.942	.	-4.994	4.635
2018m1	2	272.515	.	3.356	4.266
2018m2	2	240.836	.	4.274	4.295
2018m3	2	208.025	.	1.263	4.351
2018m4	2	204.157	.	-2.080	4.329
2018m5	2	211.528	.	-32.796	4.377
2018m6	2	202.495	.	-1.861	8.425
2018m7	2	213.607	.	4.467	8.416
2018m8	2	264.525	.	7.692	8.487
2018m9	2	249.940	.	-0.666	8.663
2018m10	2	216.116	.	1.548	8.664
2018m11	2	244.817	.	-2.075	8.648
2018m12	2	249.433	.	1.585	8.618
2019m1	2	325.179	.	-0.953	7.682
2019m2	2	329.299	.	3.043	7.681
2019m3	2	344.103	.	0.955	7.667
2019m4	2	333.627	.	8.806	7.672
2019m5	2	289.573	.	1.005	7.952
2019m6	2	310.557	.	3.172	7.959
2019m7	2	379.147	.	2.136	7.989
2019m8	2	391.923	.	5.596	8.006
2005m4	3	127.245	0	15.468	3.459
2005m5	3	123.800	0	9.475	4.256
2005m6	3	114.818	0	-4.145	4.111
2005m7	3	188.646	0	5.979	4.462
2005m8	3	196.787	0	1.820	4.453
2005m9	3	226.247	0	7.193	4.231
2005m10	3	423.601	0	5.949	4.267
2005m11	3	410.346	0	11.414	4.294
2005m12	3	410.163	0	3.415	4.582
2006m1	3	286.244	0	-3.638	4.531
2006m2	3	243.892	0	3.525	4.782
2006m3	3	315.259	0	-2.674	4.349
2006m4	3	400.426	0	-15.387	4.559
2006m5	3	365.981	0	3.361	6.112
2006m6	3	330.578	0	-3.119	6.113
2006m7	3	397.946	0	0.815	6.237
2006m8	3	372.446	0	-5.920	6.239
2006m9	3	377.925	0	2.545	6.485
2006m10	3	283.294	0	5.069	6.482
2006m11	3	293.269	0	2.837	6.504
2006m12	3	283.235	0	-7.028	6.497
2007m1	3	254.011	0	3.150	6.790
2007m2	3	239.677	0	-3.596	6.786
2007m3	3	284.697	0	0.419	6.148
2007m4	3	300.458	0	4.698	5.846
2007m5	3	341.561	0	4.674	5.812
2007m6	3	378.400	0	7.016	5.765
2007m7	3	356.199	0	5.441	5.910
2007m8	3	414.115	0	3.390	5.835
2007m9	3	373.504	0	1.079	5.758
2007m10	3	323.737	0	1.274	5.262
2007m11	3	308.236	0	-0.638	5.224
2007m12	3	341.313	0	-5.763	5.153
2008m1	3	313.278	0	4.120	5.273
2008m2	3	474.994	0	7.224	5.303
2008m3	3	484.543	0	3.708	4.171
2008m4	3	420.665	0	1.968	4.180
2008m5	3	421.700	0	2.782	4.055
2008m6	3	383.516	0	2.177	4.057
2008m7	3	358.284	0	2.899	3.678
2008m8	3	287.105	0	-3.146	3.681
2008m9	3	233.982	0	-3.608	3.784
2008m10	3	239.454	0	-11.796	3.943
2008m11	3	207.342	0	-27.122	4.520
2008m12	3	246.320	0	-9.803	7.524
2009m1	3	209.569	0	-13.137	7.772
2009m2	3	235.872	0	-14.020	8.227
2009m3	3	291.180	0	-4.845	8.568
2009m4	3	334.400	0	4.560	8.461
2009m5	3	346.066	1.190	11.362	8.348
2009m6	3	315.585	0.720	7.556	8.701
2009m7	3	311.061	0	-0.859	8.871
2009m8	3	327.981	0	-5.087	8.849
2009m9	3	350.421	0	3.012	8.839
2009m10	3	310.032	0	-1.233	8.907
2009m11	3	307.854	0.960	-6.198	8.872
2009m12	3	290.303	0.800	-5.797	8.808
2010m1	3	270.171	0	1.027	8.562
2010m2	3	246.901	0.850	0.285	8.480
2010m3	3	251.524	0.780	4.909	8.439
2010m4	3	230.153	0	2.076	8.523
2010m5	3	204.730	0.670	-3.158	8.520
2010m6	3	176.639	0.600	-8.573	8.413
2010m7	3	192.908	0	-2.364	8.484
2010m8	3	203.097	0.590	1.693	8.488
2010m9	3	195.306	0.530	0.505	8.354
2010m10	3	200.562	0	4.336	6.402
2010m11	3	198.771	0.550	1.123	6.261
2010m12	3	210.318	0.510	-2.478	5.664
2011m1	3	190.060	0	0.932	4.818
2011m2	3	175.159	0.670	-1.394	4.692
2011m3	3	182.389	0	-5.670	4.609
2011m4	3	177.288	0	-2.904	4.015
2011m5	3	163.713	0.700	3.528	3.587
2011m6	3	164.426	0.840	-2.225	3.725
2011m7	3	150.275	0	-2.645	3.623
2011m8	3	150.557	1	-2.959	3.536
2011m9	3	146.115	0.930	-4.049	3.558
2011m10	3	138.592	0	0.719	3.441
2011m11	3	132.747	0.920	1.472	3.294
2011m12	3	135.738	0	-1.841	3.307
2012m1	3	142.828	0	0.045	3.306
2012m2	3	166.873	0.720	0.940	3.054
2012m3	3	169.428	0	4.299	3.007
2012m4	3	207.998	0	0.620	3.184
2012m5	3	178.505	0.760	3.249	2.677
2012m6	3	175.387	.	-2.783	2.746
2012m7	3	199.794	.	-6.756	2.770
2012m8	3	198.437	.	-1.196	3.082
2012m9	3	186.788	.	2.455	2.874
2012m10	3	166.131	.	2.028	2.932
2012m11	3	161.935	.	-3.673	2.970
2012m12	3	174.920	.	3.023	3.015
2013m1	3	220.187	.	-0.163	3.123
2013m2	3	218.947	.	5.105	2.915
2013m3	3	211.413	.	3.437	3.083
2013m4	3	248.997	.	3.915	3.078
2013m5	3	236.498	.	10.028	3.148
2013m6	3	250.086	.	11.072	3.705
2013m7	3	225.984	.	-6.568	4.199
2013m8	3	211.463	.	3.754	4.387
2013m9	3	224.919	.	-5.575	4.414
2013m10	3	205.671	.	1.746	4.663
2013m11	3	195.289	.	2.284	4.616
2013m12	3	222.012	.	-2.046	4.611
2014m1	3	189.979	.	-3.078	4.669
2014m2	3	180.276	.	2.696	4.707
2014m3	3	192.405	.	-0.816	4.721
2014m4	3	179.375	.	-1.571	4.710
2014m5	3	171.741	.	-2.205	4.673
2014m6	3	174.767	.	-1.586	4.410
2014m7	3	190.678	.	-4.483	4.421
2014m8	3	188.317	.	2.261	4.565
2014m9	3	183.740	.	4.116	4.568
2014m10	3	165.995	.	2.538	4.507
2014m11	3	160.971	.	-3.470	4.498
2014m12	3	172.458	.	-8.631	4.598
2015m1	3	196.668	.	-3.268	4.931
2015m2	3	166.319	.	0.558	4.939
2015m3	3	178.428	.	0.443	4.865
2015m4	3	176.216	.	-4.952	4.313
2015m5	3	162.692	.	1.494	3.543
2015m6	3	167.969	.	-1.335	3.382
2015m7	3	139.665	.	-1.433	3.218
2015m8	3	153.123	.	0.815	3.063
2015m9	3	154.570	.	-7.178	3.031
2015m10	3	153.199	.	-1.639	3.211
2015m11	3	148.099	.	0.859	3.208
2015m12	3	128.959	.	0.466	3.220
2016m1	3	127.170	.	-3.280	3.124
2016m2	3	126.703	.	-9.338	3.149
2016m3	3	128.796	.	1.800	3.573
2016m4	3	117.123	.	0.409	3.651
2016m5	3	113.430	.	3.071	3.677
2016m6	3	122.965	.	0.158	3.738
2016m7	3	111.032	.	-0.664	3.672
2016m8	3	109.408	.	1.598	3.475
2016m9	3	106.934	.	-0.576	3.429
2016m10	3	111.491	.	-0.394	3.406
2016m11	3	116.664	.	0.050	3.001
2016m12	3	148.003	.	2.800	2.966
2017m1	3	128.165	.	3.427	3.054
2017m2	3	148.711	.	17.278	3.174
2017m3	3	149.019	.	-0.722	4.836
2017m4	3	135.809	.	3.567	4.835
2017m5	3	143.771	.	-2.688	4.870
2017m6	3	140.023	.	-0.846	4.901
2017m7	3	139.212	.	-0.333	4.907
2017m8	3	141.249	.	1.305	4.614

Please let me know if this data isn't helpful. If you think it is possible to copy a large sample including more IDs, please let me know.

Best Regards

[Aamina Khurraa]

Clyde Schechter

Please see the above post.

Thanks

[Clyde Schechter]

OK. -dataex- has a -count()- option. Had you specified that with a large enough number, you would have been able to get -dataex- output of a sufficient number of observations. Nevertheless, I have worked with what you show. It isn't clear how this data relates to what you describe in number 1. There is no variable named ID, nor any named x, nor y. I'm going to guess that fi_n is the ID variable and that y is l_mktr and x is xiy. With those assumptions, I cannot replicate your problem--this runs just fine for me and produces the expected output:

. * Example generated by -dataex-. To install: ssc install dataex
. clear

. input int time byte fi_n float(xiy l_mtbills l_mktr l_ev)

         time      fi_n        xiy  l_mtbills     l_mktr       l_ev
  1. 614 2 1399.921    0   -5.67 4.609
  2. 615 2 1469.539    0  -2.904 4.015
  3. 616 2 1469.539   .7   3.528 3.587
  4. 617 2 1472.498  .84  -2.225 3.725
  5. 618 2 1538.159    0  -2.645 3.623
  6. 619 2 1538.159    1  -2.959 3.536
  7. 620 2 1541.468  .93  -4.049 3.558
  8. 621 2 1615.177    0    .719 3.441
  9. 622 2 1615.177  .92   1.472 3.294
 10. 623 2 1615.177    0  -1.841 3.307
 11. 624 2  2053.13    0    .045 3.306
 12. 625 2  2053.13  .72     .94 3.054
 13. 626 2  2053.13    0   4.299 3.007
 14. 627 2 2217.788    0     .62 3.184
 15. 628 2 2217.788  .76   3.249 2.677
 16. 629 2 2217.788    .  -2.783 2.746
 17. 630 2  792.354    .  -6.756  2.77
 18. 631 2  792.354    .  -1.196 3.082
 19. 632 2  792.354    .   2.455 2.874
 20. 633 2  787.717    .   2.028 2.932
 21. 634 2  355.369    .  -3.673  2.97
 22. 635 2   654.85    .   3.023 3.015
 23. 636 2   865.75    .   -.163 3.123
 24. 637 2 1093.095    .   5.105 2.915
 25. 638 2  1073.19    .   3.437 3.083
 26. 639 2 1103.127    .   3.915 3.078
 27. 640 2  919.176    .  10.028 3.148
 28. 641 2 1017.167    .  11.072 3.705
 29. 642 2  1083.82    .  -6.568 4.199
 30. 643 2  906.396    .   3.754 4.387
 31. 644 2  936.887    .  -5.575 4.414
 32. 645 2 1109.634    .   1.746 4.663
 33. 646 2  929.986    .   2.284 4.616
 34. 647 2  969.215    .  -2.046 4.611
 35. 648 2  805.993    .  -3.078 4.669
 36. 649 2  780.052    .   2.696 4.707
 37. 650 2  761.091    .   -.816 4.721
 38. 651 2  752.463    .  -1.571  4.71
 39. 652 2   655.99    .  -2.205 4.673
 40. 653 2  589.394    .  -1.586  4.41
 41. 654 2  673.029    .  -4.483 4.421
 42. 655 2  729.991    .   2.261 4.565
 43. 656 2  720.554    .   4.116 4.568
 44. 657 2   665.44    .   2.538 4.507
 45. 658 2  598.092    .   -3.47 4.498
 46. 659 2  534.727    .  -8.631 4.598
 47. 660 2  567.922    .  -3.268 4.931
 48. 661 2  558.627    .    .558 4.939
 49. 662 2  491.337    .    .443 4.865
 50. 663 2  516.265    .  -4.952 4.313
 51. 664 2  474.058    .   1.494 3.543
 52. 665 2  465.644    .  -1.335 3.382
 53. 666 2  449.315    .  -1.433 3.218
 54. 667 2  389.505    .    .815 3.063
 55. 668 2  362.365    .  -7.178 3.031
 56. 669 2  352.028    .  -1.639 3.211
 57. 670 2  387.508    .    .859 3.208
 58. 671 2  351.798    .    .466  3.22
 59. 672 2  322.062    .   -3.28 3.124
 60. 673 2  321.063    .  -9.338 3.149
 61. 674 2  300.825    .     1.8 3.573
 62. 675 2  302.035    .    .409 3.651
 63. 676 2   282.03    .   3.071 3.677
 64. 677 2  243.787    .    .158 3.738
 65. 678 2  250.284    .   -.664 3.672
 66. 679 2  231.223    .   1.598 3.475
 67. 680 2  227.958    .   -.576 3.429
 68. 681 2   216.23    .   -.394 3.406
 69. 682 2  199.837    .     .05 3.001
 70. 683 2  240.195    .     2.8 2.966
 71. 684 2  392.783    .   3.427 3.054
 72. 685 2  348.468    .  17.278 3.174
 73. 686 2  330.129    .   -.722 4.836
 74. 687 2   348.85    .   3.567 4.835
 75. 688 2  308.723    .  -2.688  4.87
 76. 689 2  300.169    .   -.846 4.901
 77. 690 2  330.785    .   -.333 4.907
 78. 691 2  325.953    .   1.305 4.614
 79. 692 2  321.889    .    .589 4.586
 80. 693 2  293.353    .   -3.13 4.587
 81. 694 2  242.059    .  -2.515 4.663
 82. 695 2  235.942    .  -4.994 4.635
 83. 696 2  272.515    .   3.356 4.266
 84. 697 2  240.836    .   4.274 4.295
 85. 698 2  208.025    .   1.263 4.351
 86. 699 2  204.157    .   -2.08 4.329
 87. 700 2  211.528    . -32.796 4.377
 88. 701 2  202.495    .  -1.861 8.425
 89. 702 2  213.607    .   4.467 8.416
 90. 703 2  264.525    .   7.692 8.487
 91. 704 2   249.94    .   -.666 8.663
 92. 705 2  216.116    .   1.548 8.664
 93. 706 2  244.817    .  -2.075 8.648
 94. 707 2  249.433    .   1.585 8.618
 95. 708 2  325.179    .   -.953 7.682
 96. 709 2  329.299    .   3.043 7.681
 97. 710 2  344.103    .    .955 7.667
 98. 711 2  333.627    .   8.806 7.672
 99. 712 2  289.573    .   1.005 7.952
100. 713 2  310.557    .   3.172 7.959
101. 714 2  379.147    .   2.136 7.989
102. 715 2  391.923    .   5.596 8.006
103. 543 3  127.245    0  15.468 3.459
104. 544 3    123.8    0   9.475 4.256
105. 545 3  114.818    0  -4.145 4.111
106. 546 3  188.646    0   5.979 4.462
107. 547 3  196.787    0    1.82 4.453
108. 548 3  226.247    0   7.193 4.231
109. 549 3  423.601    0   5.949 4.267
110. 550 3  410.346    0  11.414 4.294
111. 551 3  410.163    0   3.415 4.582
112. 552 3  286.244    0  -3.638 4.531
113. 553 3  243.892    0   3.525 4.782
114. 554 3  315.259    0  -2.674 4.349
115. 555 3  400.426    0 -15.387 4.559
116. 556 3  365.981    0   3.361 6.112
117. 557 3  330.578    0  -3.119 6.113
118. 558 3  397.946    0    .815 6.237
119. 559 3  372.446    0   -5.92 6.239
120. 560 3  377.925    0   2.545 6.485
121. 561 3  283.294    0   5.069 6.482
122. 562 3  293.269    0   2.837 6.504
123. 563 3  283.235    0  -7.028 6.497
124. 564 3  254.011    0    3.15  6.79
125. 565 3  239.677    0  -3.596 6.786
126. 566 3  284.697    0    .419 6.148
127. 567 3  300.458    0   4.698 5.846
128. 568 3  341.561    0   4.674 5.812
129. 569 3    378.4    0   7.016 5.765
130. 570 3  356.199    0   5.441  5.91
131. 571 3  414.115    0    3.39 5.835
132. 572 3  373.504    0   1.079 5.758
133. 573 3  323.737    0   1.274 5.262
134. 574 3  308.236    0   -.638 5.224
135. 575 3  341.313    0  -5.763 5.153
136. 576 3  313.278    0    4.12 5.273
137. 577 3  474.994    0   7.224 5.303
138. 578 3  484.543    0   3.708 4.171
139. 579 3  420.665    0   1.968  4.18
140. 580 3    421.7    0   2.782 4.055
141. 581 3  383.516    0   2.177 4.057
142. 582 3  358.284    0   2.899 3.678
143. 583 3  287.105    0  -3.146 3.681
144. 584 3  233.982    0  -3.608 3.784
145. 585 3  239.454    0 -11.796 3.943
146. 586 3  207.342    0 -27.122  4.52
147. 587 3   246.32    0  -9.803 7.524
148. 588 3  209.569    0 -13.137 7.772
149. 589 3  235.872    0  -14.02 8.227
150. 590 3   291.18    0  -4.845 8.568
151. 591 3    334.4    0    4.56 8.461
152. 592 3  346.066 1.19  11.362 8.348
153. 593 3  315.585  .72   7.556 8.701
154. 594 3  311.061    0   -.859 8.871
155. 595 3  327.981    0  -5.087 8.849
156. 596 3  350.421    0   3.012 8.839
157. 597 3  310.032    0  -1.233 8.907
158. 598 3  307.854  .96  -6.198 8.872
159. 599 3  290.303   .8  -5.797 8.808
160. 600 3  270.171    0   1.027 8.562
161. 601 3  246.901  .85    .285  8.48
162. 602 3  251.524  .78   4.909 8.439
163. 603 3  230.153    0   2.076 8.523
164. 604 3   204.73  .67  -3.158  8.52
165. 605 3  176.639   .6  -8.573 8.413
166. 606 3  192.908    0  -2.364 8.484
167. 607 3  203.097  .59   1.693 8.488
168. 608 3  195.306  .53    .505 8.354
169. 609 3  200.562    0   4.336 6.402
170. 610 3  198.771  .55   1.123 6.261
171. 611 3  210.318  .51  -2.478 5.664
172. 612 3   190.06    0    .932 4.818
173. 613 3  175.159  .67  -1.394 4.692
174. 614 3  182.389    0   -5.67 4.609
175. 615 3  177.288    0  -2.904 4.015
176. 616 3  163.713   .7   3.528 3.587
177. 617 3  164.426  .84  -2.225 3.725
178. 618 3  150.275    0  -2.645 3.623
179. 619 3  150.557    1  -2.959 3.536
180. 620 3  146.115  .93  -4.049 3.558
181. 621 3  138.592    0    .719 3.441
182. 622 3  132.747  .92   1.472 3.294
183. 623 3  135.738    0  -1.841 3.307
184. 624 3  142.828    0    .045 3.306
185. 625 3  166.873  .72     .94 3.054
186. 626 3  169.428    0   4.299 3.007
187. 627 3  207.998    0     .62 3.184
188. 628 3  178.505  .76   3.249 2.677
189. 629 3  175.387    .  -2.783 2.746
190. 630 3  199.794    .  -6.756  2.77
191. 631 3  198.437    .  -1.196 3.082
192. 632 3  186.788    .   2.455 2.874
193. 633 3  166.131    .   2.028 2.932
194. 634 3  161.935    .  -3.673  2.97
195. 635 3   174.92    .   3.023 3.015
196. 636 3  220.187    .   -.163 3.123
197. 637 3  218.947    .   5.105 2.915
198. 638 3  211.413    .   3.437 3.083
199. 639 3  248.997    .   3.915 3.078
200. 640 3  236.498    .  10.028 3.148
201. 641 3  250.086    .  11.072 3.705
202. 642 3  225.984    .  -6.568 4.199
203. 643 3  211.463    .   3.754 4.387
204. 644 3  224.919    .  -5.575 4.414
205. 645 3  205.671    .   1.746 4.663
206. 646 3  195.289    .   2.284 4.616
207. 647 3  222.012    .  -2.046 4.611
208. 648 3  189.979    .  -3.078 4.669
209. 649 3  180.276    .   2.696 4.707
210. 650 3  192.405    .   -.816 4.721
211. 651 3  179.375    .  -1.571  4.71
212. 652 3  171.741    .  -2.205 4.673
213. 653 3  174.767    .  -1.586  4.41
214. 654 3  190.678    .  -4.483 4.421
215. 655 3  188.317    .   2.261 4.565
216. 656 3   183.74    .   4.116 4.568
217. 657 3  165.995    .   2.538 4.507
218. 658 3  160.971    .   -3.47 4.498
219. 659 3  172.458    .  -8.631 4.598
220. 660 3  196.668    .  -3.268 4.931
221. 661 3  166.319    .    .558 4.939
222. 662 3  178.428    .    .443 4.865
223. 663 3  176.216    .  -4.952 4.313
224. 664 3  162.692    .   1.494 3.543
225. 665 3  167.969    .  -1.335 3.382
226. 666 3  139.665    .  -1.433 3.218
227. 667 3  153.123    .    .815 3.063
228. 668 3   154.57    .  -7.178 3.031
229. 669 3  153.199    .  -1.639 3.211
230. 670 3  148.099    .    .859 3.208
231. 671 3  128.959    .    .466  3.22
232. 672 3   127.17    .   -3.28 3.124
233. 673 3  126.703    .  -9.338 3.149
234. 674 3  128.796    .     1.8 3.573
235. 675 3  117.123    .    .409 3.651
236. 676 3   113.43    .   3.071 3.677
237. 677 3  122.965    .    .158 3.738
238. 678 3  111.032    .   -.664 3.672
239. 679 3  109.408    .   1.598 3.475
240. 680 3  106.934    .   -.576 3.429
241. 681 3  111.491    .   -.394 3.406
242. 682 3  116.664    .     .05 3.001
243. 683 3  148.003    .     2.8 2.966
244. 684 3  128.165    .   3.427 3.054
245. 685 3  148.711    .  17.278 3.174
246. 686 3  149.019    .   -.722 4.836
247. 687 3  135.809    .   3.567 4.835
248. 688 3  143.771    .  -2.688  4.87
249. 689 3  140.023    .   -.846 4.901
250. 690 3  139.212    .   -.333 4.907
251. 691 3  141.249    .   1.305 4.614
252. end

. format %tm time

. 
. capture program drop one_id

. program define one_id
  1.     qreg l_mktr xiy
  2.     gen coeff = _b[xiy]
  3.     gen se = _se[xiy]
  4.     exit
  5. end

. 
. runby one_id, by(fi_n) status verbose
Iteration  1:  WLS sum of weighted deviations =  165.68546

Iteration  1: sum of abs. weighted deviations =  165.28218
Iteration  2: sum of abs. weighted deviations =  165.00504

Median regression                                   Number of obs =        102
  Raw sum of deviations    165.1 (about .40900001)
  Min sum of deviations  165.005                    Pseudo R2     =     0.0006

------------------------------------------------------------------------------
      l_mktr |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
         xiy |   .0001101   .0008328     0.13   0.895    -.0015422    .0017624
       _cons |    .375734    .727378     0.52   0.607    -1.067363    1.818831
------------------------------------------------------------------------------
Iteration  1:  WLS sum of weighted deviations =  284.41446

Iteration  1: sum of abs. weighted deviations =  284.73163
Iteration  2: sum of abs. weighted deviations =  280.75249
Iteration  3: sum of abs. weighted deviations =  280.53192
Iteration  4: sum of abs. weighted deviations =  280.52097

Median regression                                   Number of obs =        149
  Raw sum of deviations  290.349 (about .505)
  Min sum of deviations  280.521                    Pseudo R2     =     0.0338

------------------------------------------------------------------------------
      l_mktr |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
         xiy |    .010507   .0046795     2.25   0.026     .0012593    .0197547
       _cons |  -1.565437   1.137977    -1.38   0.171    -3.814346    .6834721
------------------------------------------------------------------------------

  elapsed ----------- by-groups ----------    ------- observations ------       time
     time      count     errors    no-data        processed         saved  remaining
------------------------------------------------------------------------------------
 00:00:00          2          0          0              251           251   00:00:00

--------------------------------------
Number of by-groups    =             2
by-groups with errors  =             0
by-groups with no data =             0
Observations processed =           251
Observations saved     =           251
--------------------------------------

. 
end of do-file

. list in 1/5, noobs clean

      time   fi_n        xiy   l_mtbi~s   l_mktr    l_ev      coeff         se  
    2011m3      2   1399.921          0    -5.67   4.609   .0001101   .0008328  
    2011m4      2   1469.539          0   -2.904   4.015   .0001101   .0008328  
    2011m5      2   1469.539         .7    3.528   3.587   .0001101   .0008328  
    2011m6      2   1472.498        .84   -2.225   3.725   .0001101   .0008328  
    2011m7      2   1538.159          0   -2.645   3.623   .0001101   .0008328

So I don't know what to tell you.

[Aamina Khurraa]

Clyde Schechter

Dear Clyde,

I apologies for not telling you the variables specifications.

FI_n is the panle ID, Xiy=y variable and all others are the X variables i.e. l_mtbi~s l_mktr l_ev

Please also note that this is just a small part of my data but my IDs in total are 260. I am still trying again but not getting what you have shown here.

Best
Aamina

[Clyde Schechter]

OK, what you are describing now is very different from your request in #1. You do not have a single predictor variable here, you have several. That means that the code in program one_id must be modified to store the coefficients and standard errors for each of those. I suspect that your problem is that you have not done that, or have done it incorrectly.

capture program drop one_id

program define one_id
    qreg xiy l_mtbills l_mktr l_ev
    foreach v of varlist l_mtbills l_mktr l_ev {
        gen coeff_`v' = _b[`v']
        gen se_`v' = _se[`v']
    }
    exit
end


runby one_id, by(fi_n) status

This works with your example data and produces sensible-looking results.

If you are still getting a status report indicating that all by-groups contained errors (or that a substantial number of them did), then I recommend you add the -verbose- option to your -runby- command. That won't fix anything, but it will cause Stata to show you the group-by-group output. That output will include error messages that will show you what is going wrong inside program one_id when you run it with your data, and that will enable you to make corrections to the code or to your data as needed. Once you get it working, you can then take out the -verbose- option for a final production run that will not be cluttered with the output from each by-group.

[Aamina Khurraa]

Thank you so much Clyde Schechter ,

It works perfect in generating coeff and SE for all X variables now. Just to ask, is there a way we can extract those per ID in the data, as in the data view file, the same values are repeated for each time entity of each panel as it is a time panel data with T ranging Monthly from year 2000 to 2019. The output looks something like as below and it is only for Panel ID =1.

coeff_L_MktR se_L_MktR coeff_L_EV se_L_EV
-261.4751 357.8233 5754.764 472.975
-261.4751 357.8233 5754.764 472.975
-261.4751 357.8233 5754.764 472.975
-261.4751 357.8233 5754.764 472.975
-261.4751 357.8233 5754.764 472.975
-261.4751 357.8233 5754.764 472.975
-261.4751 357.8233 5754.764 472.975
-261.4751 357.8233 5754.764 472.975
-261.4751 357.8233 5754.764 472.975
-261.4751 357.8233 5754.764 472.975
-261.4751 357.8233 5754.764 472.975
-261.4751 357.8233 5754.764 472.975

Another thing is that since it is a panel data, shouldn't I use xtqreg instead of qreg only? and do I need to mention fe(fixed effects) or re(random effects) with the regression codes here?

Also as I wan to do qreg at 0.05 lowest quantile, can I add q(0.05) to the above one_id code that you have metioned.

Thanks again in anticipation

[Clyde Schechter]

Just to ask, is there a way we can extract those per ID in the data, as in the data view file

I don't understand what this means. You showed what you have (and it is what I expected you to have). But I don't understand from your explanation what you want instead.

Another thing is that since it is a panel data, shouldn't I use xtqreg instead of qreg only? and do I need to mention fe(fixed effects) or re(random effects) with the regression codes here?

No. What -runby- does is break the data up and run -qreg- on each panel one at a time. Within each panel there is no longer any panel structure, so you use qreg, not -xtqreg-. This is more analogous to doing -xtqreg, fe- than -re-, but it isn't the same as either because in -xtqreg- you are constraining every coefficient to be the same in every ID, whereas the very purpose of this analysis is to estimate separate coefficients in each ID.

Also as I wan to do qreg at 0.05 lowest quantile, can I add q(0.05) to the above one_id code that you have metioned.

Yes.

[Aamina Khurraa]

Thanks again Clyde for a wonderful explanation.

By the first part of the questions, I tried to mean was that when Coeff and Se variables are generated in the stata data file, the same values of the coeffs are repeated for every time unit(T) of say panel ID 1(like I showed in the data I pasted before), Is it possible to extract those Coeff and Se one for each panel may be in the form of a table.

Also, I tried to run the time invariant regression where the Y variable was the returns of all the panels in the system and X variable is only one Panel returns at a time. when I run the code:

qreg Y(all panels returns) X(returns of panel=1 or 2 or 3....)

There is no output and it says convergence not achieved and failed to compute vce. Also sometimes it says data is collinear, may be because X(panel1) is also a part of Y(X all panels). Can you see how to get around this? IS that I have to use syntax "fe" with qreg

Can you also mention if there is a code to run the entire data or total panels' return (Y) regression with the individual panel returns one by one(X1, X2...). Do I need to separately generate variables as R1, R2 .... for all the individual Panel's returns(X1, X2..) and then regress them with the total panels'returns(Y) or can this be done through some code?

Best Regards

[Clyde Schechter]

By the first part of the questions, I tried to mean was that when Coeff and Se variables are generated in the stata data file, the same values of the coeffs are repeated for every time unit(T) of say panel ID 1(like I showed in the data I pasted before), Is it possible to extract those Coeff and Se one for each panel may be in the form of a table.

collapse b_* se_*, by(ID)

Also, I tried to run the time invariant regression where the Y variable was the returns of all the panels in the system and X variable is only one Panel returns at a time. when I run the code:

qreg Y(all panels returns) X(returns of panel=1 or 2 or 3....)

There is no output and it says convergence not achieved and failed to compute vce. Also sometimes it says data is collinear, may be because X(panel1) is also a part of Y(X all panels). Can you see how to get around this?

How is this new Y variable (all panel returns) calculated? Also, this would seem to require some different data structure from what you are working with. I don't really understand what you want to do here.

[Aamina Khurraa]

Thanks a million Clyde Schechter

The Y variable is not any new variable, but already existing in the dataset. It is actually the asset return of each panel for each time period.and I want to regress the whole system(all panels included) returns(Xi) with the individual panel returns(xi, for FI_n=1) where i represents individual panel(firms) in the data. Please have a look at this data below: FI_n is the panel ID and the Y variable is Xi, X variable is the Xi of the individual panel (1,2,3) at a time. That means regression would be:

qreg Xi Xi(if FI_n=1)

qreg Xi Xi(if FI_n=2) and so on for 260 Panel IDs.


Time FI_n Xi
2013m10 1 -.68829912
2013m11 1 -.65554559
2013m12 1 -.6428563
2014m1 1 -.49947822
2014m2 1 -.63309532
2014m3 1 -.53743768
2014m4 1 -.65272576
2014m5 1 -.65446103
2014m6 1 -.6711632
2014m7 1 -.70261526
2014m8 1 -.64925516
2014m9 1 -.61379027
2014m10 1 -.68233407
2014m11 1 -.65706396
2014m12 1 -.65619636
2015m1 1 -.65619636
2015m2 1 -.65619636
2015m3 1 -.65619636
2015m4 1 -.65619636
2015m5 1 -.65619636
2015m6 1 -.65619636
2015m7 1 -.65619636
2015m8 1 -.65619636
2015m9 1 -.65619636
2015m10 1 -.65619636
2015m11 1 -.65619636
2015m12 1 -.65619636
2016m1 1 -.65619636
2016m2 1 -.65619636
2016m3 1 -.65619636
2016m4 1 -.65619636
2016m5 1 -.65619636
2016m6 1 -.65619636
2016m7 1 -.65619636
2016m8 1 -.65619636
2016m9 1 -.65619636
2016m10 1 -.65619636
2016m11 1 -.65619636
2016m12 1 -.65619636
2017m1 1 -.65619636
2017m2 1 -.65619636
2017m3 1 -.53526855
2017m4 1 -.53483474
2017m5 1 -.53516006
2017m6 1 -.53516006
2017m7 1 -.53516006
2017m8 1 -.53516006
2017m9 1 -.53516006
2017m10 1 -.53516006
2017m11 1 -.53516006
2017m12 1 -.53516006
2018m1 1 -.53516006
2018m2 1 -.53516006
2018m3 1 -.53516006
2018m4 1 -.53516006
2018m5 1 -.53516006
2018m6 1 -.53516006
2018m7 1 -.53888303
2018m8 1 -.53888303
2018m9 1 -.21644539
2018m10 1 .86932051
2018m11 1 1.8060787
2018m12 1 1.8667576
2019m1 1 1.9795086
2019m2 1 1.9983363
2019m3 1 1.9778948
2019m4 1 1.9763885
2019m5 1 1.9914508
2019m6 1 1.7823019
2019m7 1 2.0497627
2019m8 1 2.2976427
2012m11 2 355.36917
2012m12 2 654.85046
2013m1 2 865.75024
2013m2 2 1093.0953
2013m3 2 1073.1897
2013m4 2 1103.1268
2013m5 2 919.17633
2013m6 2 1017.1666
2013m7 2 1083.8204
2013m8 2 906.39642
2013m9 2 936.88745
2013m10 2 1109.6343
2013m11 2 929.98602
2013m12 2 969.21454
2014m1 2 805.99261
2014m2 2 780.05219
2014m3 2 761.09082
2014m4 2 752.46313
2014m5 2 655.98969
2014m6 2 589.3941
2014m7 2 673.02863
2014m8 2 729.99103
2014m9 2 720.55396
2014m10 2 665.44006
2014m11 2 598.09229
2014m12 2 534.72717
2015m1 2 567.92194
2015m2 2 558.62708
2015m3 2 491.33698
2015m4 2 516.26483
2015m5 2 474.05835
2015m6 2 465.64438
2015m7 2 449.31503
2015m8 2 389.5047
2015m9 2 362.36499
2015m10 2 352.0278
2015m11 2 387.50766
2015m12 2 351.79791
2016m1 2 322.06165
2016m2 2 321.06311
2016m3 2 300.82455
2016m4 2 302.03534
2016m5 2 282.02972
2016m6 2 243.78677
2016m7 2 250.28363
2016m8 2 231.22293
2016m9 2 227.95763
2016m10 2 216.23006
2016m11 2 199.8372
2016m12 2 240.1955
2017m1 2 392.78323
2017m2 2 348.46765
2017m3 2 330.12857
2017m4 2 348.85031
2017m5 2 308.72275
2017m6 2 300.16937
2017m7 2 330.78461
2017m8 2 325.95261
2017m9 2 321.88885
2017m10 2 293.35306
2017m11 2 242.05925
2017m12 2 235.94232
2018m1 2 272.51468
2018m2 2 240.83572
2018m3 2 208.02515
2018m4 2 204.15651
2018m5 2 211.52809
2018m6 2 202.49466
2018m7 2 213.60698
2018m8 2 264.52454
2018m9 2 249.94032
2018m10 2 216.1156
2018m11 2 244.81712
2018m12 2 249.43275
2019m1 2 325.17941
2019m2 2 329.29874
2019m3 2 344.10272
2019m4 2 333.62732
2019m5 2 289.57346
2019m6 2 310.55719
2019m7 2 379.14719
2019m8 2 391.92252
2011m12 3 135.73845
2012m1 3 142.82813
2012m2 3 166.87296
2012m3 3 169.42847
2012m4 3 207.99799
2012m5 3 178.50516
2012m6 3 175.3868
2012m7 3 199.79358
2012m8 3 198.43716
2012m9 3 186.78813
2012m10 3 166.13101
2012m11 3 161.93526
2012m12 3 174.91965
2013m1 3 220.18668
2013m2 3 218.94662
2013m3 3 211.4133
2013m4 3 248.99706
2013m5 3 236.49826
2013m6 3 250.08646
2013m7 3 225.98436
2013m8 3 211.4628
2013m9 3 224.91881
2013m10 3 205.67104
2013m11 3 195.28877
2013m12 3 222.01204
2014m1 3 189.9785
2014m2 3 180.2758
2014m3 3 192.40457
2014m4 3 179.37495
2014m5 3 171.74084
2014m6 3 174.76704
2014m7 3 190.67757
2014m8 3 188.31688
2014m9 3 183.73953
2014m10 3 165.9951
2014m11 3 160.9707
2014m12 3 172.45792
2015m1 3 196.66788
2015m2 3 166.31879
2015m3 3 178.42822
2015m4 3 176.21552
2015m5 3 162.69159
2015m6 3 167.96936
2015m7 3 139.6649
2015m8 3 153.12274
2015m9 3 154.56984
2015m10 3 153.19868
2015m11 3 148.0993
2015m12 3 128.95886
2016m1 3 127.1698
2016m2 3 126.70283
2016m3 3 128.79564
2016m4 3 117.12334
2016m5 3 113.43034
2016m6 3 122.96464
2016m7 3 111.0322
2016m8 3 109.40765
2016m9 3 106.93421
2016m10 3 111.4911
2016m11 3 116.66447
2016m12 3 148.00323
2017m1 3 128.16495
2017m2 3 148.71143
2017m3 3 149.01871
2017m4 3 135.80881
2017m5 3 143.77136
2017m6 3 140.02286
2017m7 3 139.21159
2017m8 3 141.24895
2017m9 3 139.95776
2017m10 3 135.87836
2017m11 3 126.39616
2017m12 3 135.79721
2018m1 3 132.92706
2018m2 3 107.04241
2018m3 3 96.90815
2018m4 3 78.522408
2018m5 3 82.812103
2018m6 3 92.082253
2018m7 3 94.417168
2018m8 3 99.399483
2018m9 3 105.04055
2018m10 3 99.216026
2018m11 3 94.808296
2018m12 3 104.48966
2019m1 3 101.48612
2019m2 3 105.99895
2019m3 3 107.15462
2019m4 3 84.620842
2019m5 3 89.71418
2019m6 3 90.633171
2019m7 3 100.25871
2019m8 3 706.08984
2019m9 3 702.59308

2. Another thing that I have encountered while using one_id runby code is that some of my panels' 2, 4 and others observations are deleted altogether, although they have the data for both Y var as well as X variables in the original dataset. The resultant data in stata with new var _coeff and _se have dropped those panels altogether. please see below my runby one-id output in stata:

runby one_id, by(FI_n) status

elapsed ----------- by-groups ---------- ------- observations ------ time
time count errors no-data processed saved remaining
------------------------------------------------------------------------------------
00:00:01 2 1 0 301 153 00:02:07
00:00:02 13 5 0 1,900 1,276 00:00:39
00:00:03 25 11 0 3,907 2,359 00:00:27
00:00:04 41 17 0 6,148 3,793 00:00:21
(now reporting every 5 seconds)
00:00:09 116 41 0 16,906 12,058 00:00:11
00:00:14 191 72 0 26,935 18,930 00:00:06
00:00:19 262 93 0 38,101 27,529 00:00:00

--------------------------------------
Number of by-groups = 262
by-groups with errors = 93
by-groups with no data = 0
Observations processed = 38,101
Observations saved = 27,529
--------------------------------------

3. Lastly, I have to run another regression as:

Xit(all panels) = alpha + Xit(each panel one by one) + X2t+X3t+X4t

here again can I use runby and one_id commands to regress the entire Xit of all panles as in the data above with Xit(of individual i panels) together with other x variables.If yes, can you please specify the code?

Also how would I generate my syntax i the original one_id code to generate the constants(alphas) for each of the regression I ran before in # 9
gen coeff_`v' = _b[`v'] gen se_`v' = _se[`v'] Can I use const_'v' _a['v'] (but this doesnot work) Looking forward to you intellectual solutions to the mentioned issues as always.
Best Regards

[Clyde Schechter]

1. I'm sorry but I still don't get what you are trying to do. It looks like you want to regress Xi against itself. That makes no sense to me.

2. So something is wrong with your data. Re-run the code adding the -verbose- option to the -runby- command and then look for the error messages that come up along the way. They will tell you what the problem is.

3. Another one where I don't understand what you are trying to do. What is X2t? What is X3t, X4t? What does (each panel one by one) mean. You are writing equations that I cannot understand. How would you regress Xit from all panels against something from only one panel? I have no idea what you're talking about.

Also how would I generate my syntax i the original one_id code to generate the constants(alphas) for each of the regression I ran before in # 9
gen coeff_`v' = _b[`v'] gen se_`v' = _se[`v'] Can I use const_'v' _a['v'] (but this doesnot work)

It would be -gen constant_term = _b[_cons]- and -gen se_constant = _se[_cons]-.