model_ks: KS curve: Kolmogorov - Smirnov curve

In Dfranzani/MTRGLM: Generalized linear model metrics

KS curve: Kolmogorov - Smirnov curve

Description

model_ks() returns the Kolmogorov-Smirnov graph, which seeks to compare the cumulative distributions of successes versus failures. It is possible to choose between different graphs to study this, the first illustrates the cumulative distributions, while the second uses the true positive ratio (TPR) and the false positive ratio (FPR). The graphs allow the discrimination capacity of the model to be evaluated.

Usage

model_ks(y_train, y_train_prob, y_test, y_test_prob, type = 3)

Arguments

y_train	Values of the response variable in training set.
y_train_prob	Probabilities of the response variable in training set.
y_test	Values of the response variable in test set.
y_test_prob	Probabilities of the response variable in test set.
type	1 indicates that the KS curve is plotted using the cumulative proportions of real zeros and ones, while 2 indicates that the FPR and TPR are used. Value 3 generates both graphs (default value).

Dfranzani/MTRGLM documentation built on March 28, 2024, 1:34 a.m.