Estimation

pydimple.estimate(estimand, level=0.95, ci_type='wald', num_boot=2000)

Estimate the value of the specified node and compute a confidence interval.

Parameters:

• estimand (Node) – The node whose value is to be estimated.

• level (float, optional) – The nominal level of the reported confidence interval. Defaults to 0.95.

• ci_type (str, optional) – The type of confidence interval to compute. One of ‘bca’, ‘percentile_boot’, or ‘wald’. Defaults to ‘wald’.

• num_boot (int, optional) – The number of bootstrap replicates to use to compute the standard error. Ignored if ci_type is ‘wald’. Defaults to 2000.

Returns:

A dictionary containing the estimate of the node (‘est’), the standard error estimate (‘se’), and the confidence interval (‘ci’).

Return type:

dict

Raises:

ValueError – If an invalid ci_type is provided.

Caution

This function automatically clears the computation graph after being called!

• Index

• Search Page