Tobit in Difference in Difference

[Olive Bat]

when i run the tobit regression
tobit lexp_care year stats g_stats_yr,ll(0)
margins,dydx (*)atmeans predict (ystar(0,.))

"numerical derivatives are approximate
flat or discontinuous region encountered
numerical derivatives are approximate
flat or discontinuous region encountered
numerical derivatives are approximate
flat or discontinuous region encountered"
why do i get the above outcome mentioned in quotes ..please reply

[Phil Bromiley]

You'll increase your chances of a useful answer by following the FAQ on asking questions - provide Stata code in code delimiters, readable Stata output, and sample data using dataex. From your posting, it's not clear whether the error applies to the Tobit or to the margin statement.

Your statement seems appropriate, although I would usually leave a space before atmeans. I could be wrong, but if you doing this at the mean you're probably not getting predicted values from the xb of zero. If this is correct, then you might be able to use just the xb prediction rather than the ystar. I don't know if this would help. An alternative to the dydx would be to work strictly with predictive margins at varying values of right-hand side variables