library(testthat)
library(kernlab)

# Split the iris dataset into training and test sets
set.seed(123)  # Set seed for reproducibility
train_index <- sample(1:nrow(iris), size = 0.7 * nrow(iris))  # 70% for training
train_data <- iris[train_index, ]
test_data <- iris[-train_index, ]

test_that("svm_analysis returns a valid model", {
  # Use the training set
  formula <- Species ~ Sepal.Length + Sepal.Width

  # Call the svm_analysis function to train the model
  model <- svm_analysis(train_data, formula)

  # Check if the returned object is of class 'ksvm'
  expect_true("ksvm" %in% class(model))

  # If the model is a list, check if it contains a ksvm object
  if (is.list(model)) {
    expect_true(any(grepl("ksvm", class(model[[1]]))))
    # Check if the ksvm model has coefficients and support vectors
    expect_true(length(model[[1]]@coef) > 0)
    expect_true(model[[1]]@nSV > 0)
  } else {
    # If the model is not a list, directly check the model's structure
    expect_true(length(model@coef) > 0)
    expect_true(model@nSV > 0)
  }
})

test_that("svm_predict returns correct predictions", {
  # Use the training set to train the model
  formula <- Species ~ Sepal.Length + Sepal.Width
  model <- svm_analysis(train_data, formula)

  # Use the model to make predictions on the test set
  predictions <- svm_predict(model, test_data)

  # Check if predictions are of the expected type (factor) and length
  expect_type(predictions, "integer")
  expect_equal(length(predictions), nrow(test_data))
})

test_that("svm_plot runs without error", {
  # Use the training set
  formula <- Species ~ Sepal.Length + Sepal.Width
  model <- svm_analysis(train_data, formula)

  # Check that the plotting function runs without error
  expect_error(svm_plot(model, train_data, formula), NA)
})

test_that("svm_plot handles PCA for high-dimensional data", {
  # Use the training set with more features (high-dimensional data)
  formula <- Species ~ Sepal.Length + Sepal.Width + Petal.Length + Petal.Width
  model <- svm_analysis(train_data, formula)

  # Check that the plotting function runs without error for PCA reduction
  expect_error(svm_plot(model, train_data, formula), NA)
})