Multi-Model Performance Comparison

Compare performance of multiple statistical models side-by-side. Supports Cox proportional hazards, logistic regression, and linear regression models. Provides unified comparison tables with AIC, BIC, R², C-index, and other metrics. Inspired by Orange Data Mining's Test & Score widget, adapted for clinical research with comprehensive model diagnostics.

Usage

modelperformance(
  data,
  modelType = "cox",
  outcome = NULL,
  outcomeLevel,
  timeVar = NULL,
  model1vars = NULL,
  model1name = "Model 1",
  model2vars = NULL,
  model2name = "Model 2",
  model3vars = NULL,
  model3name = "Model 3",
  model4vars = NULL,
  model4name = "Model 4",
  model5vars = NULL,
  model5name = "Model 5",
  showAIC = TRUE,
  showRSquared = TRUE,
  showCIndex = TRUE,
  showLogLik = FALSE,
  showMCC = TRUE,
  crossValidation = FALSE,
  cvFolds = 5,
  showForestPlot = TRUE,
  showROC = FALSE,
  showCalibration = FALSE,
  autoRecommend = TRUE,
  recommendBy = "aic"
)

Arguments

data

.

modelType

.

outcome

.

outcomeLevel

.

timeVar

.

model1vars

.

model1name

.

model2vars

.

model2name

.

model3vars

.

model3name

.

model4vars

.

model4name

.

model5vars

.

model5name

.

showAIC

.

showRSquared

.

showCIndex

.

showLogLik

.

showMCC

Matthews Correlation Coefficient (MCC) - a balanced metric for binary classification, especially useful for imbalanced datasets. Ranges from -1 (total disagreement) to +1 (perfect prediction).

crossValidation

.

cvFolds

.

showForestPlot

.

showROC

.

showCalibration

.

autoRecommend

.

recommendBy

.

Value

A results object containing:

results$instructions					a html
results$comparisonTable					a table
results$cvTable					a table
results$forestPlot					an image
results$rocPlot					an image
results$calibrationPlot					an image
results$recommendation					a html
results$modelDetails					a html

Tables can be converted to data frames with asDF or as.data.frame. For example:

results$comparisonTable$asDF

as.data.frame(results$comparisonTable)

Examples

# Example: Compare Cox models
modelperformance(
    data = cancer_data,
    modelType = "cox",
    outcome = "death",
    timeVar = "months",
    model1vars = c("age", "stage"),
    model2vars = c("age", "stage", "grade"),
    model3vars = c("age", "stage", "grade", "ki67")
)