Stored results

In Stata, did_multiplegt_stat is an eclass program. In Python, the same information is available either as attributes on the fitted DIDMultiplegtStat model or as keys in the dict returned by the functional API.

Stata e()-results ↔ Python equivalents

Stata	Python (class)	Python (functional dict)
e(N)	model.n_obs_	res["results"]["N"]
e(depvar)	model._Y	res["args"]["Y"]
e(b)	model.get_coefficients()	from res["results"]["table"]
e(V) (variance matrix)	derived from model.table_["SE"]	res["results"]["table"]["SE"]

Effects' estimation

Stata	Python
e(AS)	model.to_dataframe()  or  res["results"]["table"]  — rows labelled AS, as_2, ...
e(WAS)	rows labelled WAS, was_2, ...
e(IWAS)	rows labelled IWAS, iwas_2, ...

Placebos' estimation

Stata	Python
e(Placebo_p_AS)	res["results"][f"table_placebo_{p}"]  (row 0)
e(Placebo_p_WAS)	res["results"][f"table_placebo_{p}"]  (row 1)
e(Placebo_p_IWAS)	res["results"][f"table_placebo_{p}"]  (row 2)

Or, via the class:

model.placebo_tables_   # {1: DataFrame, 2: DataFrame, ...}

By-group results

If the program is bysort-ed (Python: by=["..."]), or by_fd(K) / by_baseline(K) is specified, each level ℓ (or quantile k) gets its own block:

Stata	Python (functional dict)
e(AS_ℓ)	res[f"results_by_{j}"]["table"] (AS rows)
e(WAS_ℓ)	res[f"results_by_{j}"]["table"] (WAS rows)
e(IWAS_ℓ)	res[f"results_by_{j}"]["table"] (IWAS rows)
e(Placebo_p_AS_ℓ)	res[f"results_by_{j}"][f"table_placebo_{p}"]

res["by_levels"] lists the levels in order, so by_levels[j-1] is the value of ℓ for results block j.

Class-API attributes

After calling .fit(), the following attributes are populated on the DIDMultiplegtStat instance:

Attribute	Type	What it is
results_	dict	Full functional-API return value — the source of truth.
table_	pd.DataFrame	Main results table for the first (or only) by-group. Columns: Estimate, SE, LB CI, UB CI, Switchers, Stayers.
placebo_tables_	dict[int, pd.DataFrame] \| None	Placebo tables keyed by placebo index.
n_obs_	int	Number of observations.
n_clusters_	int \| None	Number of clusters (if cluster= was set).
by_levels_	list \| None	Levels of by-group analysis.
first_stage_	DIDMultiplegtStat \| None	Nested fitted model for the first stage of IV-WAS.
is_fitted_	bool	Whether .fit() has been called.

Notes for estout users

The Stata command is compatible with estout. In Python, the closest workflow is to pd.concat([m1.to_dataframe(), m2.to_dataframe()], keys=["m1", "m2"]) and export to LaTeX / CSV / Excel from pandas.