logistic.acc: Contingency table: Predicted vs. Observed Outcomes

In RTIInternational/mobForest: Model Based Random Forest Analysis

Description

This function takes predicted probabilities (for out of bag cases) obtained through logistic regression-based tree models and converts them into binary classes (based on specified probability threshold). The predicted classifications are then compared to actual binary response.

Usage

logistic.acc(response, predicted, prob_thresh = 0.5)

Arguments

response	A vector of binary classes of out-of-cases for a given tree.
predicted	A vector of predicted probabilities of out-of-cases using same tree.
prob_thresh	Probability threshold for classification (default = .5).

Examples

# We should get 15 predictions correct and miss 5
response <- matrix(c(rep(0,10), rep(1,10)))
predicted <- c(rep(.1,15), rep(.8,5))
logistic.acc(response, predicted, .5)	

RTIInternational/mobForest documentation built on Aug. 3, 2019, 8:28 a.m.