causalis.data_contracts.causal_diagnostic_data

Module Contents

Classes

DiagnosticData	Base class for all diagnostic data_contracts.
UnconfoundednessDiagnosticData	Fields common to all models assuming unconfoundedness.
MultiUnconfoundednessDiagnosticData	Fields common to all models assuming unconfoundedness with multi_unconfoundedness.
DiffInMeansDiagnosticData	Diagnostic data_contracts for Difference-in-Means model.
CUPEDDiagnosticData	Diagnostic data_contracts for CUPED-style (Lin-interacted OLS) adjustment.

API

class causalis.data_contracts.causal_diagnostic_data.DiagnosticData(/, **data: Any)

Bases: pydantic.BaseModel

Base class for all diagnostic data_contracts.

Initialization

Create a new model by parsing and validating input data from keyword arguments.

Raises [ValidationError][pydantic_core.ValidationError] if the input data cannot be validated to form a valid model.

self is explicitly positional-only to allow self as a field name.

model_config

‘ConfigDict(…)’

class causalis.data_contracts.causal_diagnostic_data.UnconfoundednessDiagnosticData(/, **data: Any)

Bases: causalis.data_contracts.causal_diagnostic_data.DiagnosticData

Fields common to all models assuming unconfoundedness.

Initialization

Create a new model by parsing and validating input data from keyword arguments.

Raises [ValidationError][pydantic_core.ValidationError] if the input data cannot be validated to form a valid model.

self is explicitly positional-only to allow self as a field name.

m_hat: numpy.ndarray

None

m_hat_raw: Optional[numpy.ndarray]

None

d: numpy.ndarray

None

y: Optional[numpy.ndarray]

None

x: Optional[numpy.ndarray]

None

g0_hat: Optional[numpy.ndarray]

None

g1_hat: Optional[numpy.ndarray]

None

w: Optional[numpy.ndarray]

None

w_bar: Optional[numpy.ndarray]

None

psi_b: Optional[numpy.ndarray]

None

folds: Optional[numpy.ndarray]

None

trimming_threshold: float

0.0

normalize_ipw: Optional[bool]

None

sigma2: Optional[float]

None

nu2: Optional[float]

None

psi_sigma2: Optional[numpy.ndarray]

None

psi_nu2: Optional[numpy.ndarray]

None

riesz_rep: Optional[numpy.ndarray]

None

m_alpha: Optional[numpy.ndarray]

None

psi: Optional[numpy.ndarray]

None

score: Optional[str]

None

sensitivity_analysis: Optional[Dict[str, Any]]

None

score_plot_cache: Optional[Dict[str, Any]]

None

residual_plot_cache: Optional[Dict[str, Any]]

None

class causalis.data_contracts.causal_diagnostic_data.MultiUnconfoundednessDiagnosticData(/, **data: Any)

Bases: causalis.data_contracts.causal_diagnostic_data.DiagnosticData

Fields common to all models assuming unconfoundedness with multi_unconfoundedness.

Initialization

Create a new model by parsing and validating input data from keyword arguments.

Raises [ValidationError][pydantic_core.ValidationError] if the input data cannot be validated to form a valid model.

self is explicitly positional-only to allow self as a field name.

m_hat: numpy.ndarray

None

m_hat_raw: Optional[numpy.ndarray]

None

d: numpy.ndarray

None

y: Optional[numpy.ndarray]

None

x: Optional[numpy.ndarray]

None

g_hat: Optional[numpy.ndarray]

None

psi_b: Optional[numpy.ndarray]

None

folds: Optional[numpy.ndarray]

None

trimming_threshold: float

0.0

normalize_ipw: Optional[bool]

None

sigma2: Union[float, numpy.ndarray]

None

nu2: Optional[numpy.ndarray]

None

psi_sigma2: Optional[numpy.ndarray]

None

psi_nu2: Optional[numpy.ndarray]

None

riesz_rep: Optional[numpy.ndarray]

None

m_alpha: Optional[numpy.ndarray]

None

psi: Optional[numpy.ndarray]

None

score: Optional[str]

None

sensitivity_analysis: Optional[Dict[str, Any]]

None

residual_plot_cache: Optional[Dict[str, Any]]

None

class causalis.data_contracts.causal_diagnostic_data.DiffInMeansDiagnosticData(/, **data: Any)

Bases: causalis.data_contracts.causal_diagnostic_data.DiagnosticData

Diagnostic data_contracts for Difference-in-Means model.

Initialization

Create a new model by parsing and validating input data from keyword arguments.

Raises [ValidationError][pydantic_core.ValidationError] if the input data cannot be validated to form a valid model.

self is explicitly positional-only to allow self as a field name.

class causalis.data_contracts.causal_diagnostic_data.CUPEDDiagnosticData(/, **data: Any)

Bases: causalis.data_contracts.causal_diagnostic_data.DiagnosticData

Diagnostic data_contracts for CUPED-style (Lin-interacted OLS) adjustment.

Initialization

Create a new model by parsing and validating input data from keyword arguments.

Raises [ValidationError][pydantic_core.ValidationError] if the input data cannot be validated to form a valid model.

self is explicitly positional-only to allow self as a field name.

ate_naive: float

None

se_naive: float

None

se_reduction_pct_same_cov: float

None

r2_naive: float

None

r2_adj: float

None

beta_covariates: numpy.ndarray

None

gamma_interactions: numpy.ndarray

None

covariate_outcome_corr: Optional[numpy.ndarray]

None

covariates: List[str]

None

adj_type: str

None

regression_checks: Optional[causalis.scenarios.cuped.diagnostics.regression_checks.RegressionChecks]

None