Weighting

Core weighting functions.

This module provides the main functions for applying weights to survey data using the RIM (Raking) algorithm. It includes functions to weight dataframes, return weight series, and calculate weighting efficiency metrics.

weight(df, scheme, verbose=False)

Weight a dataframe using a Rim scheme.

The dataframe must have a column for each dimension in the scheme. String columns are automatically converted to categorical, allowing easier processing.

Parameters:

Name	Type	Description	Default
df	DataFrame	The survey dataframe to weight	required
scheme	Rim	The Rim scheme object defining the weighting targets	required
verbose	bool	If True, prints progress information	False

Returns:

Type	Description
Series	A series containing the calculated weights

weight_dataframe(df, scheme, weight_column='weights', verbose=False)

Weight a dataframe using a Rim scheme and return the weighted dataframe.

The dataframe must have a column for each dimension in the scheme. String columns are automatically converted to categorical, allowing easier processing.

Parameters:

Name	Type	Description	Default
df	DataFrame	The survey dataframe to weight	required
scheme	Rim	The Rim scheme object defining the weighting targets	required
weight_column	str	Name of the column to store the calculated weights	"weights"
verbose	bool	If True, prints progress information	False

Returns:

Type	Description
DataFrame	The input dataframe with an additional weight column

weighting_efficiency(weights)

Calculate the weighting efficiency (Kish's effective sample size).

This metric indicates how much the sample size is reduced due to weighting. Higher values (closer to 100%) indicate more efficient weights.

Parameters:

Name	Type	Description	Default
weights	Series	Series containing the calculated weights	required

Returns:

Type	Description
float	The weighting efficiency as a percentage (0-100)