Add confusion matrix example to tutorial notebook

cyclops/models/plotter.py

@@ -1,152 +1,7 @@
 """Plotting functions."""
 
-import matplotlib.pyplot as plt
 import numpy as np
 import plotly.graph_objects as go
-import seaborn as sns
-
-from cyclops.models.utils import metrics_binary
-
-
-def plot_pretty_confusion_matrix(confusion_matrix: np.ndarray) -> None:
-    """Plot pretty confusion matrix.
-
-    Parameters
-    ----------
-    confusion_matrix : np.ndarray
-        confusion matrix
-
-    """
-    sns.set(style="white")
-    _, ax = plt.subplots(figsize=(9, 6))
-    sns.heatmap(
-        np.eye(2),
-        annot=confusion_matrix,
-        fmt="g",
-        annot_kws={"size": 50},
-        cmap=sns.color_palette(["tomato", "palegreen"], as_cmap=True),
-        cbar=False,
-        yticklabels=["True", "False"],
-        xticklabels=["True", "False"],
-        ax=ax,
-    )
-    ax.xaxis.tick_top()
-    ax.xaxis.set_label_position("top")
-    ax.tick_params(labelsize=20, length=0)
-
-    ax.set_title("Confusion Matrix for Test Set", size=24, pad=20)
-    ax.set_xlabel("Predicted Values", size=20)
-    ax.set_ylabel("Actual Values", size=20)
-
-    additional_texts = [
-        "(True Positive)",
-        "(False Negative)",
-        "(False Positive)",
-        "(True Negative)",
-    ]
-    for text_elt, additional_text in zip(ax.texts, additional_texts):
-        ax.text(
-            *text_elt.get_position(),
-            "\n" + additional_text,
-            color=text_elt.get_color(),
-            ha="center",
-            va="top",
-            size=24,
-        )
-    plt.tight_layout()
-    plt.show()
-
-
-def plot_confusion_matrix(confusion_matrix: np.ndarray, class_names: list) -> go.Figure:
-    """Plot confusion matrix.
-
-    Parameters
-    ----------
-    confusion_matrix : np.ndarray
-        confusion matrix
-    class_names : list
-        data class names
-
-    Returns
-    -------
-    go.Figure
-        plot figure
-
-    """
-    confusion_matrix = (
-        confusion_matrix.astype("float") / confusion_matrix.sum(axis=1)[:, np.newaxis]
-    )
-
-    layout = {
-        "title": "Confusion Matrix",
-        "xaxis": {"title": "Predicted value"},
-        "yaxis": {"title": "Real value"},
-    }
-
-    fig = go.Figure(
-        data=go.Heatmap(
-            z=confusion_matrix,
-            x=class_names,
-            y=class_names,
-            hoverongaps=False,
-            colorscale="Greens",
-        ),
-        layout=layout,
-    )
-    fig.update_layout(height=512, width=1024)
-    return fig
-
-
-def plot_auroc_across_timesteps(
-    y_pred_values: np.ndarray,
-    y_pred_labels: np.ndarray,
-    y_test_labels: np.ndarray,
-) -> go.Figure:
-    """Plot AUC_ROC across timesteps.
-
-    Parameters
-    ----------
-    y_pred_values : np.ndarray
-        prediction values
-    y_pred_labels : np.ndarray
-        prediction labels
-    y_test_labels : np.ndarray
-        data labels
-
-    Returns
-    -------
-    go.Figure
-        plot figures
-
-    """
-    num_timesteps = y_pred_labels.shape[1]
-    auroc_timesteps = []
-    for i in range(num_timesteps):
-        labels = y_test_labels[:, i]
-        pred_vals = y_pred_values[:, i]
-        preds = y_pred_labels[:, i]
-        pred_vals = pred_vals[labels != -1]
-        preds = preds[labels != -1]
-        labels = labels[labels != -1]
-        pred_metrics = metrics_binary(labels, pred_vals, preds, verbose=False)
-        auroc_timesteps.append(pred_metrics["auroc"])
-
-    print(auroc_timesteps)
-
-    prediction_hours = list(range(24, 168, 24))
-    fig = go.Figure(
-        data=[go.Bar(x=prediction_hours, y=auroc_timesteps, name="model confidence")],
-    )
-
-    fig.update_xaxes(tickvals=prediction_hours)
-    fig.update_yaxes(range=[min(auroc_timesteps) - 0.05, max(auroc_timesteps) + 0.05])
-
-    fig.update_layout(
-        title="AUROC split by no. of hours after admission",
-        autosize=False,
-        xaxis_title="No. of hours after admission",
-    )
-    return fig
 
 
 def plot_risk_mortality(predictions: np.ndarray, labels: np.ndarray) -> go.Figure:
@@ -218,13 +73,12 @@ def plot_risk_mortality(predictions: np.ndarray, labels: np.ndarray) -> go.Figur
     fig.update_yaxes(range=[label_h, 1])
     fig.update_xaxes(tickvals=prediction_hours)
     fig.update_xaxes(showline=True, linewidth=2, linecolor="black")
-
     fig.add_hline(y=0.5)
-
     fig.update_layout(
         title="Model output visualization",
         autosize=False,
         xaxis_title="No. of hours after admission",
         yaxis_title="Model confidence",
     )
+
     return fig