FactorEvaluationError fixed effects model python

Question

I'm trying to run a fixed-effects model using panel data, here is a sample of my data:

type	id	time	treatment	employment	wage
0	1	0	0	30.0	NaN
1	1	1	1	3.5	4.30
2	2	0	0	6.5	NaN
3	2	1	0	NaN	4.45
4	3	0	1	3.0	NaN

And this is the code I am using:

model_fe = PanelOLS.from_formula('employment ~ wage + EntityEffects', final_data)

However, I'm getting the error

FactorEvaluationError: Unable to evaluate factor `employment`. [NameError: `employment` is not present in the dataset or evaluation context.]

I'm unsure why this is happening, any help would be much appreciated.

Answer 1

Panel OLS usually is on muli-index, the first one being your type or id, and the second index needs to be time. So if you just set_index() your code will work:

final_data = final_data.set_index(['type', 'time'])
model_fe = PanelOLS.from_formula('employment ~ wage + EntityEffects', final_data)