Performance of Regression Models

Compute indices of model performance for regression models.

Usage

# S3 method for lm
model_performance(model, metrics = "all", verbose = TRUE, ...)

Arguments

model

A model.

metrics

Can be "all", "common" or a character vector of metrics to be computed (one or more of "AIC", "AICc", "BIC", "R2", "R2_adj", "RMSE", "SIGMA", "LOGLOSS", "PCP", "SCORE"). "common" will compute AIC, BIC, R2 and RMSE.

verbose

Toggle off warnings.

...

Arguments passed to or from other methods.

Value

A data frame (with one row) and one column per "index" (see metrics).

Details

Depending on model, following indices are computed:

• AIC Akaike's Information Criterion, see ?stats::AIC

• AICc Second-order (or small sample) AIC with a correction for small sample sizes

• BIC Bayesian Information Criterion, see ?stats::BIC

• R2 r-squared value, see r2()

• R2_adj adjusted r-squared, see r2()

• RMSE root mean squared error, see performance_rmse()

• SIGMA residual standard deviation, see insight::get_sigma()

• LOGLOSS Log-loss, see performance_logloss()

• SCORE_LOG score of logarithmic proper scoring rule, see performance_score()

• SCORE_SPHERICAL score of spherical proper scoring rule, see performance_score()

• PCP percentage of correct predictions, see performance_pcp()

model_performance() correctly detects transformed response and returns the "corrected" AIC and BIC value on the original scale. To get back to the original scale, the likelihood of the model is multiplied by the Jacobian/derivative of the transformation.

Examples

model <- lm(mpg ~ wt + cyl, data = mtcars)
model_performance(model)
#> # Indices of model performance
#> 
#> AIC     |     BIC |    R2 | R2 (adj.) |  RMSE | Sigma
#> -----------------------------------------------------
#> 156.010 | 161.873 | 0.830 |     0.819 | 2.444 | 2.568

model <- glm(vs ~ wt + mpg, data = mtcars, family = "binomial")
model_performance(model)
#> # Indices of model performance
#> 
#> AIC    |    BIC | Tjur's R2 |  RMSE | Sigma | Log_loss | Score_log | Score_spherical |   PCP
#> --------------------------------------------------------------------------------------------
#> 31.298 | 35.695 |     0.478 | 0.359 | 0.934 |    0.395 |   -14.903 |           0.095 | 0.743