r(322); Variable not found in list of covariates

[Nikita Ferguson]

Hi all,

Long-time user, first-time poster.

I am using Stata 17/SE on Mac and I am running into an r(322) error when I run margins post-Logistic regression. I've run similar code before and have not had any issues previously, so I am really unsure what is causing it.

For context, I am working with tennis data. Where 'tb' refers to a tiebreak and 'ppw' refers to the percentage of points won. Most variables are dummies, aside from a few that indicate score/number of points won/etc.

My code is as follows:

logistic set2win_t i.set1ppw_tX set1_tb_t if team_unique == 1 & set1tb_dum == 1 & set2tb_dum == 0 & gender == 0, cluster(match_id)
test i45.set1ppw_tX == i55.set1ppw_tX
levels set1ppw_tX, local(lvs)
margins, at(set1tb_ppw_t=(`lvs'))

Where set1ppw_tX is a generated int (of total percentage of points won):

gen int set1ppw_tX = round(set1tb_ppw_t, 0.05)*100

This is an example of my data:

       +-------------------------------------------------------------------------------------+
       | match_id   team_u~e   gender   set1tb~m   set2tb~m   set2wi~t   set1pp~X   set1_t~t |
       |-------------------------------------------------------------------------------------|
12817. |     3205          1        0          1          0          0         55          8 |
12818. |     3205          .        0          1          0          0         55          8 |
12819. |     3205          1        0          1          0          1         45          6 |
12820. |     3205          .        0          1          0          1         45          6 |
       |-------------------------------------------------------------------------------------|
14385. |     3597          1        1          1          0          0         30          3 |
14386. |     3597          .        1          1          0          0         30          3 |
14387. |     3597          1        1          1          0          1         70          7 |
14388. |     3597          .        1          1          0          1         70          7 |
       |-------------------------------------------------------------------------------------|
14449. |     3613          1        1          1          0          0         35          4 |
14450. |     3613          .        1          1          0          0         35          4 |
14451. |     3613          1        1          1          0          1         65          7 |
14452. |     3613          .        1          1          0          1         65          7 |
       |-------------------------------------------------------------------------------------|

The error comes when I try to run:

margins, at(set1tb_ppw_t=(`lvs'))

Am I making an obvious error here? Have I been sitting in front of Stata too long? I've run similar code before and not had any issues

Any help is appreciated!

...Also I know these variables are horribly named, haha!

[Carlo Lazzaro]

Nikita:
welcome to this forum.
Please use -dataex- to share an excerpt/example of your dataset. Thanks.

[Nikita Ferguson]

Hi Carlos,

Thanks for your reply.

* Example generated by -dataex-. For more info, type help dataex
clear
input float(match_id team_unique gender set1tb_dum set2tb_dum set2win_t) int set1ppw_tX float set1_tb_t
3205 1 0 1 0 0 55  8
3205 . 0 1 0 0 55  8
3205 1 0 1 0 1 45  6
3205 . 0 1 0 1 45  6
3597 1 1 1 0 0 30  3
3597 . 1 1 0 0 30  3
3597 1 1 1 0 1 70  7
3597 . 1 1 0 1 70  7
3613 1 1 1 0 0 35  4
3613 . 1 1 0 0 35  4
3613 1 1 1 0 1 65  7
3613 . 1 1 0 1 65  7
3628 1 1 1 0 0 40  5
3628 . 1 1 0 0 40  5
3628 1 1 1 0 1 60  7
3628 . 1 1 0 1 60  7
3629 1 1 1 0 0 35  4
3629 . 1 1 0 0 35  4
3629 1 1 1 0 1 65  7
3629 . 1 1 0 1 65  7
3660 1 1 1 0 1 45 13
3660 . 1 1 0 1 45 13
3660 1 1 1 0 0 55 15
3660 . 1 1 0 0 55 15
3661 1 1 1 0 1 20  2
3661 . 1 1 0 1 20  2
3661 1 1 1 0 0 80  7
3661 . 1 1 0 0 80  7
3677 1 1 1 0 1 90  7
3677 . 1 1 0 1 90  7
3677 1 1 1 0 0 15  1
3677 . 1 1 0 0 15  1
3692 1 1 1 0 1 70  7
3692 . 1 1 0 1 70  7
3692 1 1 1 0 0 30  3
3692 . 1 1 0 0 30  3
3700 1 1 1 0 0 35  4
3700 . 1 1 0 0 35  4
3700 1 1 1 0 1 65  7
3700 . 1 1 0 1 65  7
3708 1 1 1 0 1 40  5
3708 . 1 1 0 1 40  5
3708 1 1 1 0 0 60  7
3708 . 1 1 0 0 60  7
3710 1 1 1 0 1 60  7
3710 . 1 1 0 1 60  7
3710 1 1 1 0 0 40  5
3710 . 1 1 0 0 40  5
3726 1 1 1 0 0 40  5
3726 . 1 1 0 0 40  5
3726 1 1 1 0 1 60  7
3726 . 1 1 0 1 60  7
3731 1 1 1 0 1 20  2
3731 . 1 1 0 1 20  2
3731 1 1 1 0 0 80  7
3731 . 1 1 0 0 80  7
3738 1 1 1 0 0 30  3
3738 . 1 1 0 0 30  3
3738 1 1 1 0 1 70  7
3738 . 1 1 0 1 70  7
3739 1 1 1 0 1 80  7
3739 . 1 1 0 1 80  7
3739 1 1 1 0 0 20  2
3739 . 1 1 0 0 20  2
3741 1 1 1 0 1 70  7
3741 . 1 1 0 1 70  7
3741 1 1 1 0 0 30  3
3741 . 1 1 0 0 30  3
3745 1 1 1 0 0 70  7
3745 . 1 1 0 0 70  7
3745 1 1 1 0 1 30  3
3745 . 1 1 0 1 30  3
3751 1 1 1 0 1 20  2
3751 . 1 1 0 1 20  2
3751 1 1 1 0 0 80  7
3751 . 1 1 0 0 80  7
3769 1 1 1 0 0 70  7
3769 . 1 1 0 0 70  7
3769 1 1 1 0 1 30  3
3769 . 1 1 0 1 30  3
3770 1 1 1 0 1 70  7
3770 . 1 1 0 1 70  7
3770 1 1 1 0 0 30  3
3770 . 1 1 0 0 30  3
3777 1 1 1 0 0 45  9
3777 . 1 1 0 0 45  9
3777 1 1 1 0 1 55 11
3777 . 1 1 0 1 55 11
3786 1 1 1 0 0 80  7
3786 . 1 1 0 0 80  7
3786 1 1 1 0 1 20  2
3786 . 1 1 0 1 20  2
3796 1 1 1 0 1 45  9
3796 . 1 1 0 1 45  9
3796 1 1 1 0 0 55 11
3796 . 1 1 0 0 55 11
3801 1 1 1 0 0 55  8
3801 . 1 1 0 0 55  8
3801 1 1 1 0 1 45  6
3801 . 1 1 0 1 45  6
end

[Carlo Lazzaro]

Nikita:
the overridding issue with your example seems to rest on perfect prediction:

. logistic set2win_t i.set1ppw_tX set1_tb_t if team_unique == 1 & set1tb_dum == 1 & set2tb_dum == 0 & gender == 0, cluster(match_id)
outcome = set1ppw_tX > 0 predicts data perfectly
r(2000);

[Nikita Ferguson]

I did notice that when I run the code using the excerpt.
The logit uses 62,000 observations and the results are typical of what I would expect to see...

Here is another dataex that works when I try it:

* Example generated by -dataex-. For more info, type help dataex
clear
input float(match_id team_unique gender set1tb_dum set2tb_dum set2win_t) int set1ppw_tX float set1_tb_t
  3205 1 0 1 0 0 55  8
  3205 1 0 1 0 1 45  6
 35455 1 0 1 0 0 70  7
 35455 1 0 1 0 1 30  3
 35456 1 0 1 0 1 65  7
 35456 1 0 1 0 0 35  4
 35457 1 0 1 0 0 65  7
 35457 1 0 1 0 1 35  4
 92172 1 0 1 0 0 70  7
 92172 1 0 1 0 1 30  3
 92178 1 0 1 0 1 45  7
 92178 1 0 1 0 0 55  9
 92186 1 0 1 0 1 35  4
 92186 1 0 1 0 0 65  7
 92200 1 0 1 0 0 30  3
 92200 1 0 1 0 1 70  7
 92207 1 0 1 0 1 90  7
 92207 1 0 1 0 0 15  1
 92224 1 0 1 0 0 30  3
 92224 1 0 1 0 1 70  7
 92225 1 0 1 0 1 90  7
 92225 1 0 1 0 0 15  1
 95254 1 0 1 0 1 65  7
 95254 1 0 1 0 0 35  4
 95257 1 0 1 0 1 65  7
 95257 1 0 1 0 0 35  4
 95262 1 0 1 0 1 90  7
 95262 1 0 1 0 0 15  1
 95278 1 0 1 0 0 30  3
 95278 1 0 1 0 1 70  7
 95298 1 0 1 0 1 55  8
 95298 1 0 1 0 0 45  6
124732 1 0 1 0 0 15  1
124732 1 0 1 0 1 90  7
124736 1 0 1 0 0 80  7
124736 1 0 1 0 1 20  2
124737 1 0 1 0 1 90  7
124737 1 0 1 0 0 15  1
124738 1 0 1 0 1 20  2
124738 1 0 1 0 0 80  7
124743 1 0 1 0 0 40  5
124743 1 0 1 0 1 60  7
124746 1 0 1 0 0 60  7
124746 1 0 1 0 1 40  5
124753 1 0 1 0 0 65  7
124753 1 0 1 0 1 35  4
124761 1 0 1 0 1 40  5
124761 1 0 1 0 0 60  7
141361 1 0 1 0 1 70  7
141361 1 0 1 0 0 30  3
141374 1 0 1 0 0 60  7
141374 1 0 1 0 1 40  5
141404 1 0 1 0 1 45  6
141404 1 0 1 0 0 55  8
141412 1 0 1 0 0 20  2
141412 1 0 1 0 1 80  7
144299 1 0 1 0 0 45  8
144299 1 0 1 0 1 55 10
144301 1 0 1 0 1 20  2
144301 1 0 1 0 0 80  7
144308 1 0 1 0 1 20  2
144308 1 0 1 0 0 80  7
144310 1 0 1 0 1 65  7
144310 1 0 1 0 0 35  4
144318 1 0 1 0 1 45  7
144318 1 0 1 0 0 55  9
144322 1 0 1 0 1 60  7
144322 1 0 1 0 0 40  5
144342 1 0 1 0 . 40  5
144342 1 0 1 0 . 60  7
150763 1 0 1 0 0 70  7
150763 1 0 1 0 1 30  3
150764 1 0 1 0 1 80  7
150764 1 0 1 0 0 20  2
150775 1 0 1 0 0 55  9
150775 1 0 1 0 1 45  7
150804 1 0 1 0 1 65  7
150804 1 0 1 0 0 35  4
150811 1 0 1 0 0 45  6
150811 1 0 1 0 1 55  8
150833 1 0 1 0 0 30  3
150833 1 0 1 0 1 70  7
150869 1 0 1 0 1 40  5
150869 1 0 1 0 0 60  7
150870 1 0 1 0 1 60  7
150870 1 0 1 0 0 40  5
150880 1 0 1 0 0 40  5
150880 1 0 1 0 1 60  7
150896 1 0 1 0 0 65  7
150896 1 0 1 0 0 65  7
150915 1 0 1 0 1 70  7
150915 1 0 1 0 0 30  3
150916 1 0 1 0 0 35  4
150916 1 0 1 0 1 65  7
150921 1 0 1 0 1 40  5
150921 1 0 1 0 0 60  7
150930 1 0 1 0 0 60  7
150930 1 0 1 0 1 40  5
150932 1 0 1 0 1 30  3
150932 1 0 1 0 0 70  7
end