multiple_inference.significance_condition#

Inference for parameters that achieve statistical significance.

Classes

SignificanceCondition(*args, **kwargs)

Significance condition quantile-unbiased estimator.

SignificanceConditionResults(*args[, ...])

Quantile-unbiased results.
class multiple_inference.significance_condition.SignificanceCondition(*args: Any, **kwargs: Any)[source]#

Significance condition quantile-unbiased estimator.

Subclasses multiple_inference.base.ModelBase.

Examples

Get a quantile-unbiased distribution for x3.

import numpy as np
from multiple_inference.significance_condition import SignificanceCondition

model = SignificanceCondition(np.arange(4), np.identity(4))
dist = model.get_marginal_distribution("x3")
print(dist.ppf([.025, .5, .975]))

[-0.33936473  1.86862792  4.79906012]

Display the results.

results = model.fit()
print(results.summary(columns=["x3"]))

Significance condition quantile-unbiased estimates
===============================================
    coef (median) pvalue (1-sided) [0.025 0.975]
-----------------------------------------------
x3         1.869            0.115 -0.339  4.799
===============
Dep. Variable y
---------------
get_marginal_distribution(column: str | int, alpha: float = 0.05, **kwargs: Any) → quantile_unbiased[source]#

Get the marginal quantile-unbiased distribution.

The distribution is quantile-unbiased conditional on the parameter being statistically significant at level alpha.

Parameters:

  • column (ColumnType) – Selected column.

  • alpha (float, optional) – Significance level. Defaults to .05.

Returns:

Quantile-unbiased distribution.

Return type:

quantile_unbiased

class multiple_inference.significance_condition.SignificanceConditionResults(*args: Any, marginal_distribution_kwargs: Mapping[str, Any] | None = None, **kwargs: Any)[source]#

Quantile-unbiased results.

Sublcasses multiple_inference.base.ResultsBase.

Parameters: