Plot Neural Network Classification Landscapes

Plots landscapes with neural network classification results, highlighting correct and misclassified cases. Optionally, only misclassified landscapes can be shown.

Usage

plot_classified_landscapes(
  classification,
  landscapes,
  only_misclassified = FALSE,
  ...
)

Arguments

classification

A data frame with columns: landscape_id, actual_class, predicted_class, and confidence. Can be obtained from the CV-fold results of train_nn_metrics/train_nn_pixels or the output of apply_nn_metrics/apply_nn_pixels.

landscapes

A list of landscape objects corresponding one-to-one and in the same order as the rows in `classification`. The easiest way to ensure this is to use the same list of landscapes for both training and plotting.

only_misclassified

Logical; if TRUE, only misclassified landscapes are plotted. Default is FALSE.

...

Additional arguments passed to plot_landscape_list, such as show_legend, legend_title, ncol, max_landscapes, force, or subset_index.

Value

A patchwork object combining landscape plots with classification annotations.

See also

train_nn_pixels, train_nn_metrics

Other visualization: plot_landscape(), plot_landscape_list(), plot_metrics()

Examples

# \donttest{
# Generate training landscapes
landscapes <- create_landscapes(n = 30, patterns = c("random", "sharp", "diffuse"))
#> ✔ Successfully generated all 30 training landscapes

# Calculate landscape metrics
metrics <- calculate_landscape_metrics(landscapes, level = "landscape")
#>  ■■■■■■■■                          23% |  ETA:  9s
#>  ■■■■■■■■■■■■■                     41% |  ETA:  8s
#>  ■■■■■■■■■■■■■■■■■                 53% |  ETA:  8s
#>  ■■■■■■■■■■■■■■■■■■■■■■■■■         80% |  ETA:  3s

# Find the best 10 metrics for classification
best_10 <- evaluate_landscape_metrics(metrics, metrics_number = 10)
#> Warning: Excluded 180 rows containing 6 metrics with NA values. Metrics removed:
#> "enn_cv", "enn_mn", "enn_sd", "iji", "pafrac", and "rpr" Use
#> `exclude_NA_metrics = FALSE` to retain (not recommended for model training)
#> Warning: Excluded 3 metrics with zero variance: "pr", "prd", and "ta"
#> ! Only 6 uncorrelated metrics found. Filling to 10 with correlated metrics.

# Train model with cross-validation
model <- train_nn_metrics(metrics, metrics_selected = best_10, cv_method = "k-fold")
#> ℹ Low sample-to-predictor ratio (3:1). Consider LOO CV or reducing features.
#> ℹ Reducing CV folds from 5 to 3 to maintain 3 samples per class per fold.
#> 
#> ── Cross-validation results ──
#> 
#> ℹ Method: 3-fold cross-validation
#> ℹ Overall accuracy: 93.33%
#> 
#> ── Confusion matrix 
#>          Actual
#> Predicted diffuse random sharp
#>   diffuse       9      1     0
#>   random        1      9     0
#>   sharp         0      0    10
#> 
#> ── Per-class performance 
#> # A tibble: 3 × 5
#>   class   count recall precision f1_score
#>   <chr>   <int>  <dbl>     <dbl>    <dbl>
#> 1 diffuse    10    0.9       0.9      0.9
#> 2 random     10    0.9       0.9      0.9
#> 3 sharp      10    1         1        1  

# Plot all classification results
plot_classified_landscapes(
  model$performance$validation_results,
  landscapes
)


# Show only misclassifications without legend
plot_classified_landscapes(
  model$performance$validation_results,
  landscapes,
  only_misclassified = TRUE,
  show_legend = FALSE,
  ncol = 4
)

# }