Submodule

causalis.scenarios.did.model

model

Submodule causalis.scenarios.did.model with no child pages and 12 documented members.

Symbol index API members Estimator AggregateKind BasePeriod CallawaySantAnnaDID

Classes

Jump directly into the documented classes for this page.

1 items

CallawaySantAnnaDIDclass

Data

Jump directly into the documented data for this page.

4 items

Estimatordata AggregateKinddata BasePerioddata __all__data

data

causalis.scenarios.did.model.Estimator

Estimator

Value: None

None

Canonical target

causalis.scenarios.did.model.Estimator

Link to this symbol

data

causalis.scenarios.did.model.AggregateKind

AggregateKind

Value: None

None

Canonical target

causalis.scenarios.did.model.AggregateKind

Link to this symbol

data

causalis.scenarios.did.model.BasePeriod

BasePeriod

Value: None

None

Canonical target

causalis.scenarios.did.model.BasePeriod

Link to this symbol

class

causalis.scenarios.did.model.CallawaySantAnnaDID

CallawaySantAnnaDID

Callaway-Sant’Anna staggered-adoption DID estimator.

The model estimates group-time average treatment effects $ATT(g,t)$ for cohorts (groups) first treated in period $g$ and calendar times $t$ . Estimation is performed using the doubly robust or inverse-probability weighting methods proposed by Callaway and Sant’Anna (2021).

Parameters

estimator{“dr”, “ipw”}, default “dr”: The estimator to use for $ATT(g,t)$ cells. - “dr”: Doubly robust AIPW-style estimator. - “ipw”: Inverse probability weighting estimator.
control_group{“not_yet_or_never”, “never_treated”}, default “not_yet_or_never”: Which units to use as the comparison group. - “not_yet_or_never”: Includes units not yet treated by time $t$ and never-treated units. - “never_treated”: Includes only never-treated units.
anticipationint, default 0: Number of periods before actual treatment start where units are considered treated (e.g., due to announcement effects).
base_period{“universal”, “varying”}, default “universal”: Definition of the ‘before’ period in the DID comparison. - “universal”: Uses $g - 1 - \text{anticipation}$ as the base for all $t$ . - “varying”: Uses $t - 1$ as the base for all $t$ .
include_pre_periodsbool, default False: Whether to estimate $ATT(g,t)$ for $t < g$ (pre-treatment testing).
alphafloat, default 0.05: Significance level for confidence intervals.
diagnostic_databool, default True: Whether to store diagnostic information (overlap, balance, etc.).
propensity_clipfloat, default 1e-4: Clipping threshold for propensity scores to avoid division by zero.
logit_ridgefloat, default 1e-4: L2 regularization strength for the logistic propensity score model.
optimizer_tolfloat, default 1e-8: Tolerance for the propensity score optimizer.
optimizer_maxiterint, default 1000: Maximum iterations for the propensity score optimizer.
min_treated_per_cellint, default 30: Minimum number of treated units required in a $(g,t)$ cell.
min_control_per_cellint, default 30: Minimum number of control units required in a $(g,t)$ cell.
min_control_essfloat, default 20.0: Minimum effective sample size for control units in a $(g,t)$ cell.
max_propensity_clip_sharefloat, default 0.05: Maximum fraction of units that can be clipped before skipping a cell.
max_condition_numberfloat, default 1e5: Maximum condition number for covariate matrices.
bootstrap_replicationsint, default 0: Number of multiplier bootstrap replications for simultaneous confidence bands. If 0, uses asymptotic normal approximation for pointwise intervals.
random_stateint, optional, default None: Random seed for bootstrap multipliers.

Examples

Generate synthetic staggered-adoption panel data

Notes

In a staggered adoption design, units $i$ are treated at different times $G_i \in \{g_1, \dots, g_K, \infty\}$ . The group-time average treatment effect is defined as:

ATT(g,t) = \mathbb{E}[Y_t(g) - Y_t(\infty) \mid G=g]

where $Y_t(g)$ is the potential outcome at time $t$ if treated at time $g$ , and $Y_t(\infty)$ is the potential outcome if never treated.

This implementation follows Callaway and Sant’Anna (2021) and uses logistic regression for propensity scores and linear regression for outcome models (in “dr” mode). The doubly robust estimator for $ATT(g,t)$ is:

\theta_{DR}(g,t) = \mathbb{E} \left[ \left( \frac{G_g}{\mathbb{E}[G_g]} - \frac{\frac{p_g(X) C}{1-p_g(X)}}{\mathbb{E}[\frac{p_g(X) C}{1-p_g(X)}]} \right) (Y_t - Y_{base} - m_{g,t}(X)) \right]

where $G_g = \mathbb{1}\{G=g\}$ , $C$ is the control group indicator, $p_g(X)$ is the propensity score, and $m_{g,t}(X)$ is the outcome regression.

Simultaneous confidence bands are computed using the multiplier bootstrap on the influence functions of the $ATT(g,t)$ estimates.

Canonical target

causalis.scenarios.did.model.CallawaySantAnnaDID

Sections

ParametersNotesExamples

Link to this symbol

method

causalis.scenarios.did.model.CallawaySantAnnaDID.fit

fit

Fit all supported $ATT(g,t)$ cells on a validated :class:PanelDataDID.

This method prepares the panel data, identifies valid comparison groups for each cohort-time cell, and estimates the group-time average treatment effects.

Parameters

dataPanelDataDID: The validated panel data container.

Returns

CallawaySantAnnaDID

The fitted model instance.

Canonical target

causalis.scenarios.did.model.CallawaySantAnnaDID.fit

Sections

ParametersReturns

Link to this symbol

method

causalis.scenarios.did.model.CallawaySantAnnaDID.estimate

estimate

Aggregate group-time effects and perform inference.

This method computes simple, cohort, calendar, and event-study aggregations from the fitted $ATT(g,t)$ cells. It also performs inference using either asymptotic normal approximation or the multiplier bootstrap.

Parameters

alphafloat, optional: Significance level. If None, uses the value from __init__.
diagnostic_databool, optional: Whether to include diagnostic data in the result. If None, uses the value from __init__.
bootstrap_replicationsint, optional: Number of multiplier bootstrap replications. If None, uses the value from __init__.
random_stateint, optional: Random seed for bootstrap. If None, uses the value from __init__.

Returns

CallawaySantAnnaDIDEstimate

The estimation results including aggregations and inference.

Canonical target

causalis.scenarios.did.model.CallawaySantAnnaDID.estimate

Sections

ParametersReturns

Link to this symbol

property

causalis.scenarios.did.model.CallawaySantAnnaDID.is_fitted

is_fitted

Whether the model has been fitted on data.

Canonical target

causalis.scenarios.did.model.CallawaySantAnnaDID.is_fitted

Link to this symbol

property

causalis.scenarios.did.model.CallawaySantAnnaDID.support_

support_

A table of units and their inclusion status in various $ATT(g,t)$ cells.

Canonical target

causalis.scenarios.did.model.CallawaySantAnnaDID.support_

Link to this symbol

property

causalis.scenarios.did.model.CallawaySantAnnaDID.skipped_cells_

skipped_cells_

A table of cohort-time cells that were skipped due to insufficient data or quality issues.

Canonical target

causalis.scenarios.did.model.CallawaySantAnnaDID.skipped_cells_

Link to this symbol

property

causalis.scenarios.did.model.CallawaySantAnnaDID.cell_diagnostics_

cell_diagnostics_

Detailed diagnostic information for each $ATT(g,t)$ cell.

Canonical target

causalis.scenarios.did.model.CallawaySantAnnaDID.cell_diagnostics_

Link to this symbol

method

causalis.scenarios.did.model.CallawaySantAnnaDID.__repr__

repr

Canonical target

causalis.scenarios.did.model.CallawaySantAnnaDID.__repr__

Link to this symbol

data

causalis.scenarios.did.model.__all__

all

Value: ['AggregateKind', 'BasePeriod', 'CallawaySantAnnaDID', 'CallawaySantAnnaDIDEstimate', 'Estimator']

[‘AggregateKind’, ‘BasePeriod’, ‘CallawaySantAnnaDID’, ‘CallawaySantAnnaDIDEstimate’, ‘Estimator’]

Canonical target

causalis.scenarios.did.model.__all__

Link to this symbol