Weight of Evidence API Reference
This page documents the API for the Weight of Evidence function in Pypulate.
Weight of Evidence (WOE) and Information Value (IV) calculation.
weight_of_evidence(good_count, bad_count, min_samples=0.01, adjustment=0.5)
Calculate Weight of Evidence (WOE) and Information Value (IV).
WOE = ln(Distribution of Good / Distribution of Bad)
Parameters:
| Name | Type | Description | Default |
|---|---|---|---|
good_count
|
array_like
|
Count of good cases in each bin |
required |
bad_count
|
array_like
|
Count of bad cases in each bin |
required |
min_samples
|
float
|
Minimum percentage of samples required in a bin |
0.01
|
adjustment
|
float
|
Adjustment factor for zero counts |
0.5
|
Returns:
| Type | Description |
|---|---|
dict
|
WOE values, IV, and distributions |