regressionEvaluation: Evaluate Regression Model Results

In ivanliu1989/RQuant: A collection of generic functions

Description

Evaluate the performance of a regression model.

Usage

regressionEvaluation(obs, pred, model = NULL, sample_size = 1000,
  seed = 1234)

Arguments

obs	A vector of actual outcomes.
pred	A vector of fitted values.
model	Optional, the model used to predict fitted values from actual outcomes.
sample_size	the maximum sample size, in numbers or percents of total observations, will be used to visualise regression effects.
seed	Random seeds for reproducibility.

Value

A list of data.table and ggplot2

Examples

data(mtcars)
fit <- glm(mpg ~ cyl + disp + hp + drat, data=as.data.frame(mtcars), family = gaussian())
actuals <- mtcars$mpg
probs <- predict(fit, mtcars)
res <- regressionEvaluation(pred = probs,
obs = actuals,
model = fit,
sample_size=1e5,
seed = 1234)
res$descriptive_statistics
res$goodness_of_fit
res$model_details
res$visualisation$residual_vs_fitted
res$visualisation$residual_vs_order
res$visualisation$histograms
res$visualisation$prediction_intervals
res$check_model_assumptions

ivanliu1989/RQuant documentation built on Sept. 13, 2019, 11:53 a.m.