causalinference.core package¶
causalinference.core.data module¶
-
class
causalinference.core.data.
Dict
¶ Bases:
object
Dictionary-mimicking class.
-
class
causalinference.core.data.
Data
(outcome, treatment, covariates)¶ Bases:
causalinference.core.data.Dict
Dictionary-like class containing basic data.
causalinference.core.propensity module¶
-
class
causalinference.core.propensity.
Propensity
(data, lin, qua)¶ Bases:
causalinference.core.data.Dict
Dictionary-like class containing propensity score data.
Propensity score related data includes estimated logistic regression coefficients, maximized log-likelihood, predicted propensity scores, and lists of the linear and quadratic terms that are included in the logistic regression.
-
class
causalinference.core.propensity.
PropensitySelect
(data, lin_B, C_lin, C_qua)¶ Bases:
causalinference.core.propensity.Propensity
Dictionary-like class containing propensity score data.
Propensity score related data includes estimated logistic regression coefficients, maximized log-likelihood, predicted propensity scores, and lists of the linear and quadratic terms that are included in the logistic regression.
causalinference.core.strata module¶
-
class
causalinference.core.strata.
Strata
(strata, subsets, pscore)¶ Bases:
object
List-like object containing the stratified propensity bins.
causalinference.core.summary module¶
-
class
causalinference.core.summary.
Summary
(data)¶ Bases:
causalinference.core.data.Dict
Dictionary-like class containing summary statistics for input data.
One of the summary statistics is the normalized difference between covariates. Large values indicate that simple linear adjustment methods may not be adequate for removing biases that are associated with differences in covariates.