Skip to content

Keep never-treated units in make_did_CS2021 with float time and anticipation #437

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
SvenKlaassen merged 2 commits into from
Nov 2, 2026
Merged

Keep never-treated units in make_did_CS2021 with float time and anticipation #437

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
2003ff0
Keep never-treated units in make_did_CS2021 with float time and antic…
AsadGudzevic94 Nov 2, 2026
07a9184
update time type test
SvenKlaassen Nov 2, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
7 changes: 3 additions & 4 deletions doubleml/did/datasets/dgp_did_CS2021.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -313,16 +313,15 @@ def make_did_CS2021(n_obs=1000, dgp_type=1, include_never_treated=True, time_typ
if anticipation_periods > 0:
# filter time periods
df = df[df["t"] >= time_periods[anticipation_periods]]
# filter treatment after anticipation periods
df = df[(df["d"] <= time_periods[-(anticipation_periods + 1)]) | pd.isna(df["d"])]
# filter treatment after anticipation periods (never treated: NaT for datetime, inf for float)
never_treated = pd.isna(df["d"]) if time_type == "datetime" else np.isinf(df["d"])
df = df[(df["d"] <= time_periods[-(anticipation_periods + 1)]) | never_treated]

# update time periods by subtracting time delta
if time_type == "datetime":
df = df[(df["d"] <= time_periods[-(anticipation_periods + 1)]) | pd.isna(df["d"])]
df["t"] = df["t"].apply(lambda x: x - pd.DateOffset(months=anticipation_periods))
else:
assert time_type == "float"
df = df[(df["d"] <= time_periods[-(anticipation_periods + 1)]) | np.isinf(df["d"])]
df["t"] = df["t"] - anticipation_periods

return df
15 changes: 15 additions & 0 deletions doubleml/did/tests/test_datasets.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -72,6 +72,21 @@ def test_make_did_CS2021_return_types(dgp_type, include_never_treated, time_type
assert isinstance(df, pd.DataFrame)


@pytest.mark.ci
@pytest.mark.parametrize("anticipation_periods", [0, 1, 2])
def test_make_did_CS2021_time_types_equivalent(anticipation_periods):
"""Datetime and float time types simulate the same panel, including all never-treated units."""
np.random.seed(3141)
df_dt = make_did_CS2021(n_obs=200, time_type="datetime", anticipation_periods=anticipation_periods)
np.random.seed(3141)
df_fl = make_did_CS2021(n_obs=200, time_type="float", anticipation_periods=anticipation_periods)

pd.testing.assert_frame_equal(df_dt.drop(columns=["d", "t"]), df_fl.drop(columns=["d", "t"]))
never_treated = np.isinf(df_fl["d"])
np.testing.assert_array_equal(pd.isna(df_dt["d"]), never_treated)
assert never_treated.any()


@pytest.mark.ci
def test_make_did_CS2021_exceptions():
msg = r"time_type must be one of \('datetime', 'float'\). Got 2."
Expand Down
Loading