after refactoring predict, the predict method in sandbox.regression.gmm.IV2SLS has the wrong signature, exog and params reversed
Is this intended for 0.5.1?
We need to do something for 0.5.1.
Minimal change is to just adjust the signature of predict (to adjust to refactoring two releases ago).
But we could also (semi-silently) put my entire GMM branch into the sandbox of 0.5.1, which would also fix smaller bugs.
Or something in between.
I'd like to do 0.5.1 sometime this week/weekend, though my internet connectivity on the weekends is spotty right now. Mainly just because if we don't do it now it'll continue to slide.
Closed by #1199.