I'm trying to use Panel OLS from linearmodels, but get a message saying that the endogenous and exogenous variables aren't the same length. According to the code in the PanelOLS function, y.shape[0] != x.shape[0]
, but when I print both shapes it's the same length.
Both x
and y
are multiindex df using this format:
mi_data = data.set_index(['fcode', 'year'])
Is there something I'm missing that's particular to Panel OLS?
y variable is UPT, x is treatUberX or treatGTNotStd
Data snapshot---there's also city and state, both strings, that were not able to be captured in the shot: