Performing a log likelihood ratio test when one estimates cannot be stored

Hi,

A quick help is very much desired.

I am trying to conduct a log likelihood ratio test between two models. Stata can store estimates from one model which is built in Stata but it cannot store estimates of other model as it is a user written programme. I have outputs of both the model with Number of obs, Wald chi2(df), Prob > chi2, Log pseudolikelihood and Pseudo R2. Please help me perform the test.

Thanks very much. Ujjwal