wald_test about linearmodels HOT 4 OPEN

Hi,

Can you provide me with a copy-pastable example that raises the issue?

Thanks,
Kevin

from linearmodels.

import linearmodels as plm
from linearmodels.datasets import wage_panel

df = wage_panel.load()
df = df.set_index(["nr","year"])
idx = df.index
df = df.reset_index()
df.index=idx
fm = 'lwage ~ C(year)*educ + union + EntityEffects'
model2 = plm.PooledOLS.from_formula(formula=fm,data=df).fit(cov_type='clustered',debiased=True, cluster_entity=True)

from linearmodels.

Kevin,
Below is the copy-paste code.

Python implementation: CPython
Python version : 3.10.9
IPython version : 8.10.0

Compiler : MSC v.1916 64 bit (AMD64)
OS : Windows
Release : 10
Machine : AMD64
Processor : Intel64 Family 6 Model 154 Stepping 3, GenuineIntel
CPU cores : 20
Architecture: 64bit
linearmodels: 4.31

import linearmodels as plm
from linearmodels.datasets import wage_panel

df = wage_panel.load()
df = df.set_index(["nr","year"])
idx = df.index
df = df.reset_index()
df.index=idx
fm = 'lwage ~ C(year)*educ + union + EntityEffects'
model2 = plm.PooledOLS.from_formula(formula=fm,data=df).fit(cov_type='clustered',debiased=True, cluster_entity=True)

null = 'C(year)[T.1984]:educ = 0'
model2.wald_test(formula=null)

FormulaSyntaxError: Unknown operator '['.
C(year)⧛[⧚T.1987]:educ

null = 'C(year,Treatment(1984)):educ = 0'
model2.wald_test(formula=null)

FormulaSyntaxError: Unknown operator ':'.
C(year,Treatment(1984))⧛:⧚educ = 0

from linearmodels.

I think you need to escale terms that are not valid Python, e.g.,

null = '`C(year,Treatment(1984)):educ` = 0'

Note the ticks around the entire variable name. I'll need to check that this works, but it should.

Note: There was a recently fixed regression in formulaic that might affect escaping. This should be fixed in the next release of formulaic.

from linearmodels.