Skip to contents

Predict method for LDA

Source: R/stat_lda.R
predict.stat_lda.Rd

Predict class membership and discriminant scores for new data using a fitted LDA model.

Usage

# S3 method for class 'stat_lda'
predict(object, newdata, retain = FALSE, fold = FALSE, .collect = TRUE, ...)

Arguments

object

A stat_lda object

newdata

A tibble with the same predictor columns as training data

retain

How many LDs to return:

  • FALSE (default): No LD scores, only predictions

  • TRUE: All LDs

  • Integer (e.g., 2): First N LDs

  • Numeric 0-1 (e.g., 0.95): LDs explaining this proportion of variance

fold

How to return LD scores (only if retain != FALSE):

  • FALSE (default): Add as separate columns (LD1, LD2, ...)

  • TRUE: Fold into single list-column named "lda"

  • Character: Fold into single list-column with this name

.collect

Logical. Should predictions be added to newdata (TRUE, default) or returned as a standalone tibble (FALSE)?

...

Additional arguments (reserved)

Value

If .collect = TRUE, returns newdata with predictions added. If .collect = FALSE, returns a tibble with identifier columns (non-predictors) and predictions only. Always includes:

  • pred: Predicted class (factor)

  • prob: Posterior probability of predicted class (numeric)

  • LD1, LD2, ... : LD scores (if retain is not FALSE)

Examples

if (FALSE) { # \dontrun{
# Train LDA
lda <- boteft %>% stat_lda(type)

# Predict on new data
new_preds <- predict(lda, new_data)

# Pipe the model
new_preds <- training %>%
  stat_lda(type) %>%
  predict(testing)
} # }