cifar10-tools 0.1.0__py3-none-any.whl → 0.2.0__py3-none-any.whl

cifar10_tools/pytorch/data.py

@@ -4,7 +4,7 @@ during devcontainer creation'''
 from pathlib import Path
 from torchvision import datasets
 
-def download_cifar10_data(data_dir: str='data/pytorch/CIFAR10'):
+def download_cifar10_data(data_dir: str='data/pytorch/cifar10'):
     '''Download CIFAR-10 dataset using torchvision.datasets.'''
 
     data_dir = Path(data_dir)

cifar10_tools/pytorch/plotting.py

@@ -0,0 +1,238 @@
+'''Plotting functions for CIFAR-10 models.'''
+
+import matplotlib.pyplot as plt
+import numpy as np
+from torch.utils.data import Dataset
+
+
+def plot_sample_images(
+    dataset: Dataset,
+    class_names: list[str],
+    nrows: int = 2,
+    ncols: int = 5,
+    figsize: tuple[float, float] | None = None,
+    cmap: str = 'gray'
+) -> tuple[plt.Figure, np.ndarray]:
+    '''Plot sample images from a dataset.
+    
+    Args:
+        dataset: PyTorch dataset containing (image, label) tuples.
+        class_names: List of class names for labeling.
+        nrows: Number of rows in the grid.
+        ncols: Number of columns in the grid.
+        figsize: Figure size (width, height). Defaults to (ncols*1.5, nrows*1.5).
+        cmap: Colormap for displaying images.
+        
+    Returns:
+        Tuple of (figure, axes array).
+    '''
+    if figsize is None:
+        figsize = (ncols * 1.5, nrows * 1.5)
+    
+    fig, axes = plt.subplots(nrows=nrows, ncols=ncols, figsize=figsize)
+    axes = axes.flatten()
+
+    for i, ax in enumerate(axes):
+        # Get image and label from dataset
+        img, label = dataset[i]
+        
+        # Unnormalize and squeeze for plotting
+        img = img * 0.5 + 0.5
+        img = img.numpy().squeeze()
+        ax.set_title(class_names[label])
+        ax.imshow(img, cmap=cmap)
+        ax.axis('off')
+
+    plt.tight_layout()
+    
+    return fig, axes
+
+
+def plot_learning_curves(
+    history: dict[str, list[float]],
+    figsize: tuple[float, float] = (10, 4)
+) -> tuple[plt.Figure, np.ndarray]:
+    '''Plot training and validation loss and accuracy curves.
+    
+    Args:
+        history: Dictionary containing 'train_loss', 'val_loss', 
+                 'train_accuracy', and 'val_accuracy' lists.
+        figsize: Figure size (width, height).
+        
+    Returns:
+        Tuple of (figure, axes array).
+    '''
+    fig, axes = plt.subplots(1, 2, figsize=figsize)
+
+    axes[0].set_title('Loss')
+    axes[0].plot(history['train_loss'], label='Train')
+    axes[0].plot(history['val_loss'], label='Validation')
+    axes[0].set_xlabel('Epoch')
+    axes[0].set_ylabel('Loss (cross-entropy)')
+    axes[0].legend(loc='best')
+
+    axes[1].set_title('Accuracy')
+    axes[1].plot(history['train_accuracy'], label='Train')
+    axes[1].plot(history['val_accuracy'], label='Validation')
+    axes[1].set_xlabel('Epoch')
+    axes[1].set_ylabel('Accuracy (%)')
+    axes[1].legend(loc='best')
+
+    plt.tight_layout()
+    
+    return fig, axes
+
+
+def plot_confusion_matrix(
+    true_labels: np.ndarray,
+    predictions: np.ndarray,
+    class_names: list[str],
+    figsize: tuple[float, float] = (8, 8),
+    cmap: str = 'Blues'
+) -> tuple[plt.Figure, plt.Axes]:
+    '''Plot a confusion matrix heatmap.
+    
+    Args:
+        true_labels: Array of true class labels.
+        predictions: Array of predicted class labels.
+        class_names: List of class names for labeling.
+        figsize: Figure size (width, height).
+        cmap: Colormap for the heatmap.
+        
+    Returns:
+        Tuple of (figure, axes).
+    '''
+    from sklearn.metrics import confusion_matrix
+    
+    cm = confusion_matrix(true_labels, predictions)
+    
+    fig, ax = plt.subplots(figsize=figsize)
+
+    ax.set_title('Confusion matrix')
+    im = ax.imshow(cm, cmap=cmap)
+
+    # Add labels
+    ax.set_xticks(range(len(class_names)))
+    ax.set_yticks(range(len(class_names)))
+    ax.set_xticklabels(class_names, rotation=45, ha='right')
+    ax.set_yticklabels(class_names)
+    ax.set_xlabel('Predicted label')
+    ax.set_ylabel('True label')
+
+    # Add text annotations
+    for i in range(len(class_names)):
+        for j in range(len(class_names)):
+            color = 'white' if cm[i, j] > cm.max() / 2 else 'black'
+            ax.text(j, i, str(cm[i, j]), ha='center', va='center', color=color)
+
+    plt.tight_layout()
+    
+    return fig, ax
+
+
+def plot_class_probability_distributions(
+    all_probs: np.ndarray,
+    class_names: list[str],
+    nrows: int = 2,
+    ncols: int = 5,
+    figsize: tuple[float, float] = (12, 4),
+    bins: int = 50,
+    color: str = 'black'
+) -> tuple[plt.Figure, np.ndarray]:
+    '''Plot predicted probability distributions for each class.
+    
+    Args:
+        all_probs: Array of shape (n_samples, n_classes) with predicted probabilities.
+        class_names: List of class names for labeling.
+        nrows: Number of rows in the subplot grid.
+        ncols: Number of columns in the subplot grid.
+        figsize: Figure size (width, height).
+        bins: Number of histogram bins.
+        color: Histogram bar color.
+        
+    Returns:
+        Tuple of (figure, axes array).
+    '''
+    fig, axes = plt.subplots(nrows=nrows, ncols=ncols, figsize=figsize)
+
+    fig.suptitle('Predicted probability distributions by class', fontsize=14, y=1.02)
+    fig.supxlabel('Predicted probability', fontsize=12)
+    fig.supylabel('Count', fontsize=12)
+
+    axes = axes.flatten()
+
+    for i, (ax, class_name) in enumerate(zip(axes, class_names)):
+        # Get probabilities for this class across all samples
+        class_probs = all_probs[:, i]
+        
+        # Plot histogram
+        ax.hist(class_probs, bins=bins, color=color)
+        ax.set_title(class_name)
+        ax.set_xlim(0, 1)
+
+    plt.tight_layout()
+    
+    return fig, axes
+
+
+def plot_evaluation_curves(
+    true_labels: np.ndarray,
+    all_probs: np.ndarray,
+    class_names: list[str],
+    figsize: tuple[float, float] = (12, 5)
+) -> tuple[plt.Figure, tuple[plt.Axes, plt.Axes]]:
+    '''Plot ROC and Precision-Recall curves for multi-class classification.
+    
+    Args:
+        true_labels: Array of true class labels.
+        all_probs: Array of shape (n_samples, n_classes) with predicted probabilities.
+        class_names: List of class names for labeling.
+        figsize: Figure size (width, height).
+        
+    Returns:
+        Tuple of (figure, (ax1, ax2)).
+    '''
+    from sklearn.metrics import roc_curve, auc, precision_recall_curve, average_precision_score
+    from sklearn.preprocessing import label_binarize
+    
+    # Binarize true labels for one-vs-rest evaluation
+    y_test_bin = label_binarize(true_labels, classes=range(len(class_names)))
+
+    # Create figure with ROC and PR curves side by side
+    fig, (ax1, ax2) = plt.subplots(1, 2, figsize=figsize)
+
+    # Plot ROC curves for each class
+    ax1.set_title('ROC curves (one-vs-rest)')
+
+    for i, class_name in enumerate(class_names):
+        fpr, tpr, _ = roc_curve(y_test_bin[:, i], all_probs[:, i])
+        roc_auc = auc(fpr, tpr)
+        ax1.plot(fpr, tpr, label=class_name)
+
+    ax1.plot([0, 1], [0, 1], 'k--', label='Random classifier')
+    ax1.set_xlabel('False positive rate')
+    ax1.set_ylabel('True positive rate')
+    ax1.legend(loc='lower right', fontsize=12)
+    ax1.set_xlim([0, 1])
+    ax1.set_ylim([0, 1.05])
+
+    # Plot Precision-Recall curves for each class
+    ax2.set_title('Precision-recall curves (one-vs-rest)')
+
+    for i, class_name in enumerate(class_names):
+        precision, recall, _ = precision_recall_curve(y_test_bin[:, i], all_probs[:, i])
+        ap = average_precision_score(y_test_bin[:, i], all_probs[:, i])
+        ax2.plot(recall, precision)
+
+    # Random classifier baseline (horizontal line at class prevalence = 1/num_classes)
+    baseline = 1 / len(class_names)
+    ax2.axhline(y=baseline, color='k', linestyle='--')
+
+    ax2.set_xlabel('Recall')
+    ax2.set_ylabel('Precision')
+    ax2.set_xlim([0, 1])
+    ax2.set_ylim([0, 1.05])
+
+    plt.tight_layout()
+    
+    return fig, (ax1, ax2)

{cifar10_tools-0.1.0.dist-info → cifar10_tools-0.2.0.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: cifar10_tools
-Version: 0.1.0
+Version: 0.2.0
 Summary: Tools for training neural networks on the CIFAR-10 task with PyTorch and TensorFlow
 License: GPLv3
 License-File: LICENSE

{cifar10_tools-0.1.0.dist-info → cifar10_tools-0.2.0.dist-info}/RECORD

@@ -1,10 +1,11 @@
 cifar10_tools/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 cifar10_tools/pytorch/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-cifar10_tools/pytorch/data.py,sha256=zEDdRbcCHehDg5mdOGDopKT-uCRTjF27Q_UYTAPVEhQ,626
+cifar10_tools/pytorch/data.py,sha256=09zodpjto0xLq95tDAyq57CFh6MSYRuUBPcMmQcyKZM,626
 cifar10_tools/pytorch/evaluation.py,sha256=i4tRYOqWATVqQVkWT_fATWRbzo9ziX2DDkXKPaiQlFE,923
+cifar10_tools/pytorch/plotting.py,sha256=B1ifJxbSEDpInnVk9c3o1fjVx534TPPKTWM5iusyzrE,7494
 cifar10_tools/pytorch/training.py,sha256=Sg6NlBT_DTyLzf-Ls3bYI8-8AwGFJblRj0MDnUmGP3Q,2642
 cifar10_tools/tensorflow/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-cifar10_tools-0.1.0.dist-info/METADATA,sha256=3wdozzaT9e9M6Tf5c7EbiJ8XVXewrJiXTHxGQxhMJ0Q,1580
-cifar10_tools-0.1.0.dist-info/WHEEL,sha256=kJCRJT_g0adfAJzTx2GUMmS80rTJIVHRCfG0DQgLq3o,88
-cifar10_tools-0.1.0.dist-info/licenses/LICENSE,sha256=wtHfRwmCF5-_XUmYwrBKwJkGipvHVmh7GXJOKKeOe2U,1073
-cifar10_tools-0.1.0.dist-info/RECORD,,
+cifar10_tools-0.2.0.dist-info/METADATA,sha256=3s6_5lP8rAnEu5F9r5YKU-EqUi9UO3mNUFK1ikVgUfc,1580
+cifar10_tools-0.2.0.dist-info/WHEEL,sha256=kJCRJT_g0adfAJzTx2GUMmS80rTJIVHRCfG0DQgLq3o,88
+cifar10_tools-0.2.0.dist-info/licenses/LICENSE,sha256=wtHfRwmCF5-_XUmYwrBKwJkGipvHVmh7GXJOKKeOe2U,1073
+cifar10_tools-0.2.0.dist-info/RECORD,,