logit regression absorbing fixed ffects

[Paula de Souza Leao Spinola]

I have a dataset at the birth level with information on date of birth, hospital code and type of delivery, among other variables. The dataset comprises of several million births having taken place in ~1,000 different hospitals over 14 calendar years.

I ran the OLS regression below of the dummy for c-section (d_pc) on the dummy for bank holiday (d_hol) controlling for time fixed effects (year, month, weekday) as well as hospital fixed effects (id_hosp), which I absorbed due to the large number of hospitals. I also clustered the stardard errors at the hospital level.

reg d_pc d_hol i.year i.month i.weekday, absorb(id_hosp) vce(cluster id_hosp)

I would like to run an equivalent logit model. I tried using -clogit- as below but it did not work given the large number of hospitals. Below is the error message I got.

. clogit d_pc d_hol i.year i.month i.weekday, group(id_hosp) vce(cluster id_hosp)
note: multiple positive outcomes within groups encountered.
note: 86 groups (1,372 obs) dropped because of all positive or
      all negative outcomes.
9,811 (group size) take 8,906 (# positives) combinations results in numeric overflow; computations cannot proceed
r(1400);

I do not know whether I could use -xtlogit- as this is not a panel data per se as each row corresponds to a different birth.

Could anyone advice me on how to run a logit model while absorbing group fixed effects?

Many thanks

[Andrew Musau]

Could anyone advice me on how to run a logit model while absorbing group fixed effects?

You are not absorbing the fixed effects with conditional logit. You are conditioning them out of the likelihood.

I do not know whether I could use -xtlogit- as this is not a panel data per se as each row corresponds to a different birth.

xtlogit calls clogit, so they are the same command. A few things that I notice are that you include weekday, month and year dummies. As months are in years and days are in weeks, these sets of dummies will be collinear. So you need to include only the lowest level (days). If there are many observations within each hospital, you may include these as dummies - but I suspect that you will run into the numeric overflow problem. Stata MP allows a larger matrix, so that may be the only viable route at the end.

[Paula de Souza Leao Spinola]

Many thanks Andrew Musau!

I am aware that conditional logit doesn't absorb the fixed effects. My goal is to be able to run a logit model in which I control for multiple fixed effects. As I don't care for the coefficient estimates of the dummy fixed effects and because Stata does not allow me to run the model while including all fixed effects, I was interested in knowing if there is a way I can absorb them. PS: I have Stata MP and still was not able to run the regression (i.e. see error message in the message I posted above).

Would you have any recommendation of how I can go forward?

PS: In my model, months refer to month alone (January, February, .... December) and weekday refers to day of the week (Monday, Tuesday, ... Sunday) - thus, it is ok to include year, month and day of the week fixed effects.
In any case, I am mostly interested in being able to estimate the model while controling for the ~1,000 hospital dummies (ideally absorbing them as it is not feasible to estimate them all with Stata)

Thanks again!

[Andrew Musau]

As clogit cannot handle the size of your dataset, if you have many observations per hospital (say at least 50), then the other option is to estimate

logit d_pc d_hol i.year i.month i.weekday i. id_hosp

[Joro Kolev]

Andrew says that -xtlogit- calls -clogit-, and looking at the code of -xtlogit- this appears to be true. So the below would probably not make a difference.

But if you want to try -xtlogit-, do

xtset id_hosp

and then -xtlogit-.

[Paula de Souza Leao Spinola]

Many thanks Andrew Musau and Joro Kolev.

-clogit- and -xtlogit- do not work given the large number of hospitals.

. clogit d_pc d_hol i.year i.month i.weekday, group(id_hosp)
note: multiple positive outcomes within groups encountered.
note: 86 groups (1,372 obs) dropped because of all positive or
      all negative outcomes.
10,163 (group size) take 9,484 (# positives) combinations results in numeric overflow; computations cannot proceed
r(1400);

end of do-file

r(1400);

. do "C:\Users\paula\AppData\Local\Temp\STD43dc_000000.tmp"

. xtset id_hosp
       panel variable:  id_hosp (unbalanced)

. xtlogit d_pc d_hol i.year i.month i.weekday, fe
note: multiple positive outcomes within groups encountered.
note: 86 groups (1,372 obs) dropped because of all positive or
      all negative outcomes.
10,163 (group size) take 9,484 (# positives) combinations results in numeric overflow; computations cannot proceed
r(1400);

end of do-file

r(1400);

.

-logit- works but seems very inefficient given that I don't care about the hospital dummy coefficients. From your comments, I guess I have no other option, right?

. logit d_pc d_hol i.year i.month i.weekday i.id_hosp

note: 749.id_hosp != 0 predicts success perfectly
      749.id_hosp dropped and 10 obs not used

note: 789.id_hosp != 0 predicts success perfectly
      789.id_hosp dropped and 33 obs not used

note: 1151.id_hosp != 0 predicts success perfectly
      1151.id_hosp dropped and 12 obs not used

note: 1292.id_hosp != 0 predicts failure perfectly
      1292.id_hosp dropped and 1 obs not used

note: 1321.id_hosp != 0 predicts success perfectly
      1321.id_hosp dropped and 6 obs not used

note: 1326.id_hosp != 0 predicts success perfectly
      1326.id_hosp dropped and 2 obs not used

note: 1339.id_hosp != 0 predicts success perfectly
      1339.id_hosp dropped and 2 obs not used

note: 1411.id_hosp != 0 predicts success perfectly
      1411.id_hosp dropped and 26 obs not used

note: 1418.id_hosp != 0 predicts success perfectly
      1418.id_hosp dropped and 16 obs not used

note: 1425.id_hosp != 0 predicts success perfectly
      1425.id_hosp dropped and 6 obs not used

note: 1436.id_hosp != 0 predicts success perfectly
      1436.id_hosp dropped and 1 obs not used

note: 1459.id_hosp != 0 predicts success perfectly
      1459.id_hosp dropped and 7 obs not used

note: 1489.id_hosp != 0 predicts success perfectly
      1489.id_hosp dropped and 6 obs not used

note: 1767.id_hosp != 0 predicts success perfectly
      1767.id_hosp dropped and 4 obs not used

note: 1794.id_hosp != 0 predicts success perfectly
      1794.id_hosp dropped and 1 obs not used

note: 2646.id_hosp != 0 predicts success perfectly
      2646.id_hosp dropped and 1 obs not used

note: 2674.id_hosp != 0 predicts success perfectly
      2674.id_hosp dropped and 3 obs not used

note: 2737.id_hosp != 0 predicts success perfectly
      2737.id_hosp dropped and 1 obs not used

note: 3411.id_hosp != 0 predicts success perfectly
      3411.id_hosp dropped and 7 obs not used

note: 4461.id_hosp != 0 predicts failure perfectly
      4461.id_hosp dropped and 135 obs not used

note: 4529.id_hosp != 0 predicts success perfectly
      4529.id_hosp dropped and 10 obs not used

note: 4805.id_hosp != 0 predicts success perfectly
      4805.id_hosp dropped and 21 obs not used

note: 4819.id_hosp != 0 predicts success perfectly
      4819.id_hosp dropped and 44 obs not used

note: 5138.id_hosp != 0 predicts success perfectly
      5138.id_hosp dropped and 2 obs not used

note: 5512.id_hosp != 0 predicts success perfectly
      5512.id_hosp dropped and 1 obs not used

note: 5993.id_hosp != 0 predicts success perfectly
      5993.id_hosp dropped and 1 obs not used

note: 5994.id_hosp != 0 predicts success perfectly
      5994.id_hosp dropped and 241 obs not used

note: 5998.id_hosp != 0 predicts success perfectly
      5998.id_hosp dropped and 40 obs not used

note: 6024.id_hosp != 0 predicts success perfectly
      6024.id_hosp dropped and 7 obs not used

note: 6029.id_hosp != 0 predicts success perfectly
      6029.id_hosp dropped and 1 obs not used

note: 6041.id_hosp != 0 predicts success perfectly
      6041.id_hosp dropped and 29 obs not used

note: 6314.id_hosp != 0 predicts success perfectly
      6314.id_hosp dropped and 4 obs not used

note: 7057.id_hosp != 0 predicts success perfectly
      7057.id_hosp dropped and 4 obs not used

note: 7095.id_hosp != 0 predicts success perfectly
      7095.id_hosp dropped and 1 obs not used

note: 7471.id_hosp != 0 predicts success perfectly
      7471.id_hosp dropped and 1 obs not used

note: 8010.id_hosp != 0 predicts success perfectly
      8010.id_hosp dropped and 33 obs not used

note: 8413.id_hosp != 0 predicts success perfectly
      8413.id_hosp dropped and 1 obs not used

note: 8626.id_hosp != 0 predicts success perfectly
      8626.id_hosp dropped and 25 obs not used

note: 8791.id_hosp != 0 predicts failure perfectly
      8791.id_hosp dropped and 16 obs not used

note: 8949.id_hosp != 0 predicts success perfectly
      8949.id_hosp dropped and 14 obs not used

note: 8994.id_hosp != 0 predicts success perfectly
      8994.id_hosp dropped and 2 obs not used

note: 9027.id_hosp != 0 predicts failure perfectly
      9027.id_hosp dropped and 1 obs not used

note: 9078.id_hosp != 0 predicts success perfectly
      9078.id_hosp dropped and 1 obs not used

note: 9079.id_hosp != 0 predicts success perfectly
      9079.id_hosp dropped and 6 obs not used

note: 9136.id_hosp != 0 predicts success perfectly
      9136.id_hosp dropped and 2 obs not used

note: 9172.id_hosp != 0 predicts failure perfectly
      9172.id_hosp dropped and 1 obs not used

note: 9180.id_hosp != 0 predicts success perfectly
      9180.id_hosp dropped and 5 obs not used

note: 9193.id_hosp != 0 predicts success perfectly
      9193.id_hosp dropped and 8 obs not used

note: 9210.id_hosp != 0 predicts success perfectly
      9210.id_hosp dropped and 2 obs not used

note: 9225.id_hosp != 0 predicts success perfectly
      9225.id_hosp dropped and 4 obs not used

note: 9235.id_hosp != 0 predicts success perfectly
      9235.id_hosp dropped and 27 obs not used

note: 9246.id_hosp != 0 predicts success perfectly
      9246.id_hosp dropped and 74 obs not used

note: 9261.id_hosp != 0 predicts failure perfectly
      9261.id_hosp dropped and 16 obs not used

note: 9274.id_hosp != 0 predicts success perfectly
      9274.id_hosp dropped and 3 obs not used

note: 9333.id_hosp != 0 predicts success perfectly
      9333.id_hosp dropped and 1 obs not used

note: 9342.id_hosp != 0 predicts success perfectly
      9342.id_hosp dropped and 4 obs not used

note: 9344.id_hosp != 0 predicts success perfectly
      9344.id_hosp dropped and 3 obs not used

note: 9595.id_hosp != 0 predicts success perfectly
      9595.id_hosp dropped and 7 obs not used

note: 9634.id_hosp != 0 predicts success perfectly
      9634.id_hosp dropped and 2 obs not used

note: 10178.id_hosp != 0 predicts success perfectly
      10178.id_hosp dropped and 8 obs not used

note: 10209.id_hosp != 0 predicts success perfectly
      10209.id_hosp dropped and 9 obs not used

note: 10239.id_hosp != 0 predicts success perfectly
      10239.id_hosp dropped and 8 obs not used

note: 10284.id_hosp != 0 predicts success perfectly
      10284.id_hosp dropped and 3 obs not used

note: 10363.id_hosp != 0 predicts success perfectly
      10363.id_hosp dropped and 1 obs not used

note: 10367.id_hosp != 0 predicts success perfectly
      10367.id_hosp dropped and 9 obs not used

note: 10371.id_hosp != 0 predicts success perfectly
      10371.id_hosp dropped and 14 obs not used

note: 10396.id_hosp != 0 predicts success perfectly
      10396.id_hosp dropped and 5 obs not used

note: 10445.id_hosp != 0 predicts success perfectly
      10445.id_hosp dropped and 7 obs not used

note: 10470.id_hosp != 0 predicts success perfectly
      10470.id_hosp dropped and 3 obs not used

note: 10473.id_hosp != 0 predicts success perfectly
      10473.id_hosp dropped and 9 obs not used

note: 10620.id_hosp != 0 predicts success perfectly
      10620.id_hosp dropped and 14 obs not used

note: 10622.id_hosp != 0 predicts success perfectly
      10622.id_hosp dropped and 108 obs not used

note: 10702.id_hosp != 0 predicts success perfectly
      10702.id_hosp dropped and 3 obs not used

note: 10766.id_hosp != 0 predicts success perfectly
      10766.id_hosp dropped and 10 obs not used

note: 11066.id_hosp != 0 predicts success perfectly
      11066.id_hosp dropped and 5 obs not used

note: 11092.id_hosp != 0 predicts success perfectly
      11092.id_hosp dropped and 3 obs not used

note: 11328.id_hosp != 0 predicts failure perfectly
      11328.id_hosp dropped and 1 obs not used

note: 11332.id_hosp != 0 predicts failure perfectly
      11332.id_hosp dropped and 6 obs not used

note: 11426.id_hosp != 0 predicts success perfectly
      11426.id_hosp dropped and 5 obs not used

note: 11487.id_hosp != 0 predicts success perfectly
      11487.id_hosp dropped and 27 obs not used

note: 11501.id_hosp != 0 predicts success perfectly
      11501.id_hosp dropped and 1 obs not used

note: 11504.id_hosp != 0 predicts success perfectly
      11504.id_hosp dropped and 77 obs not used

note: 11586.id_hosp != 0 predicts success perfectly
      11586.id_hosp dropped and 5 obs not used

note: 11623.id_hosp != 0 predicts success perfectly
      11623.id_hosp dropped and 43 obs not used

note: 12037.id_hosp != 0 predicts success perfectly
      12037.id_hosp dropped and 4 obs not used

note: 12160.id_hosp != 0 predicts success perfectly
      12160.id_hosp dropped and 39 obs not used

Iteration 0:   log likelihood =  -667620.6  
Iteration 1:   log likelihood = -608881.63  
Iteration 2:   log likelihood = -598348.59  
Iteration 3:   log likelihood = -598023.51  
Iteration 4:   log likelihood = -598013.86  
Iteration 5:   log likelihood = -598008.28  
Iteration 6:   log likelihood = -598008.24  
Iteration 7:   log likelihood = -598008.24  

Logistic regression                             Number of obs     =  1,585,048
                                                LR chi2(811)      =  139224.71
                                                Prob > chi2       =     0.0000
Log likelihood = -598008.24                     Pseudo R2         =     0.1043

------------------------------------------------------------------------------
        d_pc |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
       d_hol |  -.1238542   .0258036    -4.80   0.000    -.1744283     -.07328
             |
        year |
       2007  |   .0629948   .0197972     3.18   0.001      .024193    .1017967
       2008  |   .0241893   .0193655     1.25   0.212    -.0137664    .0621449
       2009  |   .0343011   .0192082     1.79   0.074    -.0033463    .0719486
       2010  |   .1833617   .0190269     9.64   0.000     .1460696    .2206538
       2011  |    .229111   .0183719    12.47   0.000     .1931027    .2651192
       2012  |   .2595279   .0177649    14.61   0.000     .2247093    .2943464
       2013  |   .2702904    .017659    15.31   0.000     .2356794    .3049013
       2014  |   .1908104   .0175094    10.90   0.000     .1564926    .2251282
       2015  |  -.0561694   .0173002    -3.25   0.001    -.0900771   -.0222617
       2016  |  -.0459171   .0175546    -2.62   0.009    -.0803234   -.0115108
       2017  |  -.0513048   .0173643    -2.95   0.003    -.0853382   -.0172714
       2018  |  -.0649837    .017379    -3.74   0.000    -.0990458   -.0309215
       2019  |  -.0508228   .0176636    -2.88   0.004    -.0854428   -.0162028
             |
       month |
          2  |  -.0280777   .0116162    -2.42   0.016    -.0508451   -.0053104
          3  |  -.0020992   .0113306    -0.19   0.853    -.0243068    .0201084
          4  |  -.0274121   .0114612    -2.39   0.017    -.0498756   -.0049486
          5  |  -.0188333   .0113845    -1.65   0.098    -.0411466      .00348
          6  |  -.0237337   .0115172    -2.06   0.039     -.046307   -.0011604
          7  |  -.0236274   .0114905    -2.06   0.040    -.0461483   -.0011065
          8  |  -.0197022   .0115831    -1.70   0.089    -.0424046    .0030003
          9  |   -.021721   .0115309    -1.88   0.060    -.0443212    .0008791
         10  |   .0219509   .0116954     1.88   0.061    -.0009715    .0448734
         11  |  -.0142945    .011706    -1.22   0.222    -.0372379    .0086489
         12  |   .0054226   .0115614     0.47   0.639    -.0172374    .0280825
             |
     weekday |
        Mon  |   .6566597   .0091247    71.97   0.000     .6387756    .6745438
        Tue  |   .6132953   .0091802    66.81   0.000     .5953025    .6312881
        Wed  |   .6241835   .0091968    67.87   0.000     .6061581     .642209
        Thu  |   .5905933   .0091904    64.26   0.000     .5725805    .6086062
        Fri  |   .5918517   .0093095    63.58   0.000     .5736055     .610098
        Sat  |    .308055   .0094646    32.55   0.000     .2895047    .3266054
             |
     id_hosp |
        177  |  -.8834312   .0993284    -8.89   0.000    -1.078111   -.6887511
        184  |  -.3454638   .0532151    -6.49   0.000    -.4497635   -.2411641
        189  |  -1.084048   .0435779   -24.88   0.000    -1.169459   -.9986366
        194  |    -.92577   .6411533    -1.44   0.149    -2.182407    .3308674
        263  |   -2.61546   .1345733   -19.44   0.000    -2.879219   -2.351701
        270  |   -.701618   .7624748    -0.92   0.357    -2.196041    .7928053
        316  |  -.7972923    .064841   -12.30   0.000    -.9243784   -.6702063
        431  |   .4303928   .1687642     2.55   0.011     .0996211    .7611645
        433  |  -.7743769   .0658007   -11.77   0.000    -.9033438     -.64541
        434  |   .9811979   .5088168     1.93   0.054    -.0160646     1.97846
        449  |  -2.958744    .348967    -8.48   0.000    -3.642707   -2.274782
        453  |   -1.23737   .5658791    -2.19   0.029    -2.346472   -.1282669
        455  |  -1.420057   .0423718   -33.51   0.000    -1.503104    -1.33701
        457  |   -1.74464   .0428086   -40.75   0.000    -1.828543   -1.660737
        459  |   -1.55766   .0488247   -31.90   0.000    -1.653355   -1.461965
        461  |  -1.336201   .0667241   -20.03   0.000    -1.466977   -1.205424
        531  |    .445176   .2593083     1.72   0.086     -.063059     .953411
        537  |    .057504   .0772958     0.74   0.457     -.093993    .2090009
        540  |  -.9263454   .0552735   -16.76   0.000    -1.034679   -.8180114
        613  |  -.4781573    .196695    -2.43   0.015    -.8636724   -.0926422
        653  |  -1.008878   .0578289   -17.45   0.000     -1.12222   -.8955353
        667  |  -1.047735   .0437591   -23.94   0.000    -1.133502    -.961969
        673  |  -1.358087   .0466309   -29.12   0.000    -1.449482   -1.266692
        674  |  -.5882563   .0603835    -9.74   0.000    -.7066058   -.4699069
        682  |    .291795   .0582215     5.01   0.000      .177683    .4059071
        739  |  -.9650411   .1803877    -5.35   0.000    -1.318594   -.6114877
        740  |  -.3652403   .0716754    -5.10   0.000    -.5057215   -.2247591
        744  |  -.1647989   .3329501    -0.49   0.621     -.817369    .4877713
        749  |          0  (empty)
        755  |  -1.030818   .0681931   -15.12   0.000    -1.164474   -.8971621
        773  |  -1.563832   .0748153   -20.90   0.000    -1.710468   -1.417197
        786  |  -1.199292   .0505928   -23.70   0.000    -1.298452   -1.100132
        788  |   -.687755   .2427422    -2.83   0.005    -1.163521   -.2119889
        789  |          0  (empty)
        952  |  -2.054624   .0587187   -34.99   0.000     -2.16971   -1.939537
        958  |   1.250256   1.009887     1.24   0.216    -.7290873    3.229599
        996  |   -1.56549   .1745663    -8.97   0.000    -1.907634   -1.223347
        998  |  -2.056954   .0495864   -41.48   0.000    -2.154141   -1.959766
       1038  |  -1.947553   .6026693    -3.23   0.001    -3.128764   -.7663433
       1104  |  -1.966138   .3058908    -6.43   0.000    -2.565673   -1.366603
       1124  |  -.1926043   .4290509    -0.45   0.653    -1.033529      .64832
       1127  |  -2.547321   .0456083   -55.85   0.000    -2.636711    -2.45793
       1129  |  -.8503001   .0555629   -15.30   0.000    -.9592013   -.7413988
       1134  |   1.100281   .1757812     6.26   0.000     .7557561    1.444806
       1136  |  -2.461694   .0463637   -53.10   0.000    -2.552565   -2.370823
       1138  |  -1.342017    .804276    -1.67   0.095    -2.918369    .2343346
       1139  |  -1.408286   .0892929   -15.77   0.000    -1.583297   -1.233275
       1146  |  -.6224763   .1011006    -6.16   0.000    -.8206298   -.4243229
       1147  |  -.9788687   .0801392   -12.21   0.000    -1.135939   -.8217987
       1148  |  -.8993124    .111899    -8.04   0.000     -1.11863   -.6799945
       1151  |          0  (empty)
       1162  |  -1.600353   .0472431   -33.87   0.000    -1.692948   -1.507758
       1163  |  -.7027152   .0550184   -12.77   0.000    -.8105494   -.5948811
       1165  |  -1.060796   .0931122   -11.39   0.000    -1.243293   -.8782995
       1167  |  -.4038077   .0595313    -6.78   0.000    -.5204869   -.2871284
       1168  |   .5385194    .342634     1.57   0.116     -.133031     1.21007
       1170  |  -.3922354   .3325499    -1.18   0.238    -1.044021    .2595505
       1172  |  -1.862838   .0582061   -32.00   0.000     -1.97692   -1.748756
       1173  |  -1.225766   .0483409   -25.36   0.000    -1.320512   -1.131019
       1174  |  -1.916552   .0519302   -36.91   0.000    -2.018333   -1.814771
       1175  |  -2.218294   .0493389   -44.96   0.000    -2.314996   -2.121591
       1177  |  -1.078784   .0501628   -21.51   0.000    -1.177101   -.9804667
       1180  |  -.9649359   .1359932    -7.10   0.000    -1.231478   -.6983941
       1181  |   .4500814    .126329     3.56   0.000     .2024811    .6976817
       1185  |  -.9823293   .0781028   -12.58   0.000    -1.135408   -.8292507
       1186  |   .0442544   .2528712     0.18   0.861    -.4513641    .5398728
       1216  |  -.6519734   .6187345    -1.05   0.292    -1.864671     .560724
       1220  |  -.5277711   .0435582   -12.12   0.000    -.6131435   -.4423986
       1223  |  -1.594204   .5298993    -3.01   0.003    -2.632787   -.5556202
       1225  |  -.5872256   .0645219    -9.10   0.000    -.7136862    -.460765
       1227  |  -.2620193   .7443842    -0.35   0.725    -1.720986    1.196947
       1230  |  -.7279041    .406444    -1.79   0.073     -1.52452    .0687116
       1240  |  -1.515716   .0512658   -29.57   0.000    -1.616196   -1.415237
       1247  |  -1.800224   .0539187   -33.39   0.000    -1.905903   -1.694545
       1256  |  -1.061608   .0634953   -16.72   0.000    -1.186057   -.9371598
       1258  |   1.201548   .4537028     2.65   0.008     .3123063    2.090789
       1260  |  -1.118044   .0687357   -16.27   0.000    -1.252763   -.9833242
       1266  |   -.548078   .1051423    -5.21   0.000     -.754153   -.3420029
       1279  |   .8277158   .4571105     1.81   0.070    -.0682043    1.723636
       1291  |  -.4558528   .1310911    -3.48   0.001    -.7127867   -.1989189
       1292  |          0  (empty)
       1296  |  -1.003147   .0842456   -11.91   0.000    -1.168266   -.8380292
       1298  |  -.3946117   .0725657    -5.44   0.000    -.5368379   -.2523855
       1299  |  -.9006755   .1143809    -7.87   0.000    -1.124858   -.6764931
       1301  |  -1.988238   .1263687   -15.73   0.000    -2.235916    -1.74056
       1303  |  -1.583067   .0629415   -25.15   0.000     -1.70643   -1.459704
       1312  |   -.896627   .1829671    -4.90   0.000    -1.255236   -.5380181
       1313  |  -.2074857   .1241922    -1.67   0.095    -.4508978    .0359265
       1314  |  -.8532067   .0942675    -9.05   0.000    -1.037968   -.6684457
       1316  |  -1.424191   .0548175   -25.98   0.000    -1.531631   -1.316751
       1320  |  -.4161965    .203353    -2.05   0.041    -.8147611   -.0176319
       1321  |          0  (empty)
       1323  |    .629898   .2757395     2.28   0.022     .0894584    1.170338
       1326  |          0  (empty)
       1339  |          0  (empty)
       1353  |  -.5023634   .1043609    -4.81   0.000    -.7069069   -.2978199
       1354  |  -1.403346   .0455659   -30.80   0.000    -1.492653   -1.314038
       1372  |  -.8562881   .0826239   -10.36   0.000    -1.018228   -.6943483
       1375  |  -.1151647   .0879968    -1.31   0.191    -.2876353    .0573058
       1376  |  -.2774763   .2057169    -1.35   0.177     -.680674    .1257214
       1379  |  -1.036491    .082912   -12.50   0.000    -1.198996   -.8739866
       1384  |  -2.365206    .056631   -41.77   0.000    -2.476201   -2.254211
       1394  |  -.6836463   .0767698    -8.91   0.000    -.8341123   -.5331803
       1396  |  -.2435566   .1296061    -1.88   0.060    -.4975799    .0104667
       1401  |  -1.245979   .1673149    -7.45   0.000    -1.573911   -.9180484
       1411  |          0  (empty)
       1418  |          0  (empty)
       1423  |  -1.091759   .1014133   -10.77   0.000    -1.290526    -.892993
       1425  |          0  (empty)
       1436  |          0  (empty)
       1439  |  -1.815199   .2087075    -8.70   0.000    -2.224258    -1.40614
       1442  |  -1.826538   .0568553   -32.13   0.000    -1.937972   -1.715104
       1448  |  -1.502488   .1449544   -10.37   0.000    -1.786593   -1.218383
       1452  |  -.3716549    .063758    -5.83   0.000    -.4966182   -.2466916
       1453  |  -2.714367   .8250424    -3.29   0.001    -4.331421   -1.097314
       1456  |   -2.15205   .0679773   -31.66   0.000    -2.285283   -2.018817
       1459  |          0  (empty)
       1472  |  -1.830349   .0859684   -21.29   0.000    -1.998844   -1.661853
       1475  |  -.4950714   .0999683    -4.95   0.000    -.6910057   -.2991371
       1477  |   .2832298   .4221034     0.67   0.502    -.5440776    1.110537
       1481  |  -.0343685   .4238637    -0.08   0.935    -.8651261     .796389
       1484  |  -.2189997   .2129023    -1.03   0.304    -.6362806    .1982811
       1486  |  -1.885171   .0443502   -42.51   0.000    -1.972096   -1.798246
       1489  |          0  (empty)
       1498  |   .3825586   .7241755     0.53   0.597    -1.036799    1.801916
       1557  |    -2.0449   .1290383   -15.85   0.000     -2.29781   -1.791989
       1558  |  -.6607105   .0994307    -6.64   0.000    -.8555912   -.4658299
       1560  |  -1.524027   .1384903   -11.00   0.000    -1.795463   -1.252591
       1561  |  -1.160208   .0463389   -25.04   0.000    -1.251031   -1.069386
       1562  |   .1674391   .4627153     0.36   0.717    -.7394661    1.074344
       1564  |  -1.466147   .5079389    -2.89   0.004    -2.461689   -.4706056
       1565  |  -.9854506   .0490971   -20.07   0.000    -1.081679   -.8892222
       1619  |   .5486691   .4188924     1.31   0.190    -.2723449    1.369683
       1632  |   .2929911    .165495     1.77   0.077    -.0313732    .6173553
       1765  |  -.3120347   .0952043    -3.28   0.001    -.4986317   -.1254376
       1766  |   .0171501   .3016602     0.06   0.955    -.5740931    .6083933
       1767  |          0  (empty)
       1787  |  -1.302626   .0436718   -29.83   0.000    -1.388221   -1.217031
       1794  |          0  (empty)
       1818  |  -1.891834   1.228222    -1.54   0.123    -4.299105     .515437
       1819  |  -.9901599   .1079034    -9.18   0.000    -1.201647   -.7786732
       1825  |   .2002334   .1554732     1.29   0.198    -.1044884    .5049552
       1858  |   .1137883   .1867418     0.61   0.542    -.2522189    .4797956
       1889  |  -2.139529   .0453616   -47.17   0.000    -2.228436   -2.050622
       1921  |   .8047145   .7191079     1.12   0.263     -.604711     2.21414
       1953  |   -1.13609   .0598692   -18.98   0.000    -1.253431   -1.018748
       1979  |  -1.182595   .7943964    -1.49   0.137    -2.739583    .3743936
       2033  |   .8866185   .2842026     3.12   0.002     .3295916    1.443645
       2070  |   .6097913   .1063414     5.73   0.000      .401366    .8182166
       2072  |   .3986925   .2381658     1.67   0.094    -.0681038    .8654889
       2074  |   .6393971   .1818039     3.52   0.000      .283068    .9957262
       2075  |   .2925968   .0813622     3.60   0.000     .1331297    .4520638
       2079  |   .7704362   1.015453     0.76   0.448    -1.219815    2.760687
       2094  |   .6033607   .7217012     0.84   0.403    -.8111476    2.017869
       2116  |  -.8166303   .0811699   -10.06   0.000    -.9757203   -.6575403
       2117  |   .8290435   .2308452     3.59   0.000     .3765951    1.281492
       2157  |  -1.129033   .1714698    -6.58   0.000    -1.465107   -.7929582
       2254  |   -2.03817   .6311258    -3.23   0.001    -3.275154    -.801186
       2309  |  -.5721956   .1620302    -3.53   0.000    -.8897689   -.2546223
       2403  |  -.1875956   .1622316    -1.16   0.248    -.5055636    .1303725
       2449  |   .1491549   .1894671     0.79   0.431    -.2221938    .5205035
       2459  |  -.0258739   .1681634    -0.15   0.878    -.3554681    .3037204
       2460  |  -.3439512   .2112498    -1.63   0.103    -.7579932    .0700909

etc
.

[Joro Kolev]

No, I do not think that you have any other options. And what does it mean inefficient? You managed to fit the -logit- model, and the -logit- model is appropriate here for the reasons that Andrew alluded to (that you have many observations per hospital). It seems that everything looks great.

If for reporting purposes you are annoyed that you get too many estimated parameters, fit the logit model quietly, and then use the user contributed -esttab- to display your results. -esttab- has keep and drop options, so you can control which estimates it includes in the table and which it drops.

[Paula de Souza Leao Spinola]

Awesome, many thanks Joro Kolev!

[David Ochieng]

I am trying to run a mixed logit model for a data set with 6 choice sets,3 alternatives and 5 attributes.I have tried the mixlogit command but i keep getting some error messages

cmmixlogit Choice, casevars(landsz Inclevels) random(NutritionlevelModerate)
note: 138 cases (2484 obs) dropped due to no positive outcome, multiple positive outcomes, or a single
observation per case
variable Alternative has replicate levels for one or more cases; this is not allowed

Kindly advice

[Carlo Lazzaro]

David:
welcome to this forum.
Please start a new thread, as your query has a very tenous relationship with this one. Thanks.

[Qiuyi Wang]

I have met similar problems when running xtmogit. Due to a large number of school fixed effects, neighborhood fixed effects, I could not get the regression results because it is always showing that "fitting comparison model..." Is there any alternative method I could use ?

Many thanks

[Carlo Lazzaro]

Qiuyi:
the usual recipe is to start with a more parsimonious model, progressively increasing its complexity and check when problems come alive.