"Predicts failure perfectly" (Probit, logit with interaction terms)

Dear Statalist,

I am trying to run probit and logit models, but when I include interaction terms I get the following message/error:

Code:

. probit $ylist $xlist_interactions_included

note: 1.Female#1.arab != 0 predicts failure perfectly
      1.Female#1.arab dropped and 2 obs not used

note: 1.Female#1.black != 0 predicts failure perfectly
      1.Female#1.black dropped and 1 obs not used

note: 0.Female#1.indian != 0 predicts success perfectly
      0.Female#1.indian dropped and 2 obs not used

note: 1.Female#1.latino != 0 predicts failure perfectly
      1.Female#1.latino dropped and 3 obs not used

note: 1.Female#1.indian omitted because of collinearity

Here is the list of independent variables:

Code:

global xlist_interactions_included largeuniversity PrestigiousUniversity SpecializedUniversity universityinIThotspot universitytechdegree techjobsb4bc c.age_centered c.age_centered#arab c.age_centered#asian c.age_centered#black c.age_centered#indian c.age_centered#latino Female##arab Female##asian Female##black Female##indian Female##latino

I have read this https://www.statalist.org/forums/for...lure-perfectly several times and based on that discussion I tried adding .ib2 infront of Female but the "error" persists.

Any tips on how to run the model and show the results for all interaction terms?
Thank you,
Scott