Suest for comparing Non linear regressions but show "estimation sample of the model saved under model1 could not be restored"

[Alex Potter]

Hello all,

I am using suest to compare the results of Non linear regressions,

use NLS_min1, clear
keep if US==0
nl (LN_Q = {b0} + ln(1+ {A}*A1_min1+ {B}*B_min1)+{b1}*FirmID+{y1}*year_dummy1+{y2}*year_dummy2),  variables(LN_Q A_min1 B_min1 FirmID year_dummy1 year_dummy2) 
est store model1

use NLS_min1, clear
keep if US==1
nl (LN_Q = {b0} + ln(1+ {A}*A1_min1+ {B}*B_min1)+{b1}*FirmID+{y1}*year_dummy1+{y2}*year_dummy2),  variables(LN_Q A_min1 B_min1 FirmID year_dummy1 year_dummy2) 
est store model2

suest model1 model2, vce(cluster FirmID)
testnl  [model1_mean = model2_mean]: LN_Q

However, I am getting the following error instead of results:

suest model1 model2, vce(cluster FirmID)
estimation sample of the model saved under model1 could not be restored

I have searched many web pages, but I have not been able to solve this problem
Could you please give me some advice?
Many thanks in advance.

[George Ford]

Not sure suest is feasible after nl since scores aren't available.

See this for an alternative:

https://www.stata.com/statalist/archive/2010-12/msg00660.html

[George Ford]

https://www.stata.com/support/faqs/statistics/nonlinear-models-using-ml/