linearmodels icon indicating copy to clipboard operation
linearmodels copied to clipboard

Can you run fixed effects or set `check_rank=False` in `IV2SLS`?

Open ArturoSbr opened this issue 4 years ago • 1 comments

I have a panel with multiple time periods per entity. I would like to run IV2SLS with entity effects as well as other effects.

Normally, in PanelOLS, I would set entity_effects=True and other_effects=df['other_effect'] to use the fixed effects that I need. However, these parameters do not seem to be present in IV2SLS. How can this be done?

If this has not been implemented yet, is there a way to set check_rank=False like you can in a PanelOLS estimation?

I tried to create my own fixed effects by hand using pandas.get_dummies. However, the estimation stops because one of my matrices does not have full rank.

ArturoSbr avatar Nov 06 '21 16:11 ArturoSbr

You could probably purch the FE from your Y, Z and X, and then run standard IV on the purged data..

bashtage avatar Nov 11 '21 23:11 bashtage