Create and save a confusion matrix plot and table

This function takes model predictions and generates a confusion matrix. It saves the confusion matrix as a PNG figure and as a CSV table in the specified output directory, and returns the ggplot object used to create the plot.

Usage

create_confusion_outputs(predictions, model_name, output_path)

Arguments

predictions

A data frame of predicted classes returned by train_and_predict(), including a .pred_class column and a class column representing true labels.

model_name

A string representing the model name (used for naming the output files).

output_path

A string specifying the directory to save the confusion matrix plot and CSV table.

Value

A ggplot object representing the confusion matrix.

Examples

# Example usage of the `create_confusion_outputs` function
library(testthat)
library(parsnip)
library(recipes)
#> Loading required package: dplyr
#> 
#> Attaching package: 'dplyr'
#> The following object is masked from 'package:testthat':
#> 
#>     matches
#> The following objects are masked from 'package:stats':
#> 
#>     filter, lag
#> The following objects are masked from 'package:base':
#> 
#>     intersect, setdiff, setequal, union
#> 
#> Attaching package: 'recipes'
#> The following object is masked from 'package:stats':
#> 
#>     step
library(workflows)
library(dplyr)
library(rsample)
#> 
#> Attaching package: 'rsample'
#> The following object is masked from 'package:testthat':
#> 
#>     matches
set.seed(123)
data <- tibble(
  class = sample(c("A", "B"), 150, replace = TRUE),
  feature1 = rnorm(150),
  feature2 = rnorm(150)
)

# Split the data into training and testing sets
data_split <- initial_split(data, prop = 0.8)
data_train <- training(data_split)
data_test <- testing(data_split)

# Define the model specification
model_spec <- decision_tree() %>%
  set_engine("rpart") %>%
  set_mode("classification")

# Define the recipe (preprocessing)
recipe <- recipe(class ~ feature1 + feature2, data = data_train) %>%
  step_normalize(all_numeric())

# Generate predictions
predictions <- train_and_predict(model_spec, data_train, data_test, recipe)

# Convert predicted and true class labels to factors with the same levels
predictions$.pred_class <- factor(predictions$.pred_class, levels = c("A", "B"))
predictions$class <- factor(predictions$class, levels = c("A", "B"))

# Specify output path for saving the confusion matrix
output_path <- tempdir()  # Using tempdir for testing

# Create and save the confusion matrix plot and table
conf_plot <- create_confusion_outputs(predictions, "decision_tree", output_path)
print(conf_plot)