Initial notebook

Author: ruivieira

examples/TSICE.ipynb

@@ -0,0 +1,144 @@
+import matplotlib.pyplot as plt
+import pandas as pd
+from sklearn.linear_model import LinearRegression
+import numpy as np
+
+def add_timeseries(ax, ts, color="green", name="time series", showlegend=False):
+    timestamps = ts.index
+    ax.plot(timestamps, ts[ts.columns[0]], color=color, label=(name if showlegend else '_nolegend_'))
+
+def plot_timeseries(ts, color="green", ax=None, name="time series"):
+    showlegend = True
+    if type(ts) == dict:
+        data = ts
+        if type(color) == str:
+            color = {k: color for k in data}
+    elif type(ts) == list:
+        data = {}
+        for k, ts_data in enumerate(ts):
+            data[k] = ts_data
+        if type(color) == str:
+            color = {k: color for k in data}
+    else:
+        data = {}
+        data["default"] = ts
+        color = {"default": color}
+
+    if ax is None:
+        fig, ax = plt.subplots()
+
+    first = True
+    for key, ts in data.items():
+        if not first:
+            showlegend = False
+
+        add_timeseries(ax, ts, color=color[key], showlegend=showlegend, name=name)
+        first = False
+
+    return ax
+
+def plot_tsice_explanation(explanation, forecast_horizon):
+    original_ts = pd.DataFrame(explanation["data_x"])
+    perturbations = explanation["perturbations"]
+    forecasts_on_perturbations = explanation["forecasts_on_perturbations"]
+
+    new_perturbations = []
+    new_timestamps = []
+    pred_ts = []
+
+    original_ts.index.freq = pd.infer_freq(original_ts.index)
+    for i in range(1, forecast_horizon + 1):
+        new_timestamps.append(original_ts.index[-1] + (i * original_ts.index.freq))
+
+    for perturbation in perturbations:
+        new_perturbations.append(pd.DataFrame(perturbation))
+
+    for forecast in forecasts_on_perturbations:
+        pred_ts.append(pd.DataFrame(forecast, index=new_timestamps))
+
+    pred_original_ts = pd.DataFrame(
+        explanation["current_forecast"], index=new_timestamps
+    )
+
+    fig, ax = plt.subplots()
+
+    # Plot perturbed time series
+    ax = plot_timeseries(new_perturbations, color="lightgreen", ax=ax, name="perturbed timeseries samples")
+
+    # Plot original time series
+    ax = plot_timeseries(original_ts, color="green", ax=ax, name="input/original timeseries")
+
+    # Plot varying forecast range
+    ax = plot_timeseries(pred_ts, color="lightblue", ax=ax, name="forecast on perturbed samples")
+
+    # Plot original forecast
+    ax = plot_timeseries(pred_original_ts, color="blue", ax=ax, name="original forecast")
+
+    # Set labels and title
+    ax.set_xlabel("Month/Year")
+    ax.set_ylabel("sunspots")
+    ax.set_title("Time Series Individual Conditional Expectation (TSICE) Plot")
+
+    ax.legend()
+
+    # Display the plot
+    plt.show()
+
+    # Return the figure
+    return fig
+
+
+def plot_tsice_with_observed_features(explanation, feature_per_row=2):
+    df = pd.DataFrame(explanation["data_x"])
+    n_row = int(np.ceil(len(explanation["feature_names"]) / feature_per_row))
+    feat_values = np.array(explanation["feature_values"])
+
+    fig, axs = plt.subplots(n_row, feature_per_row, figsize=(15, 15))
+    axs = axs.ravel()  # Flatten the axs to iterate over it
+
+    for i, feat in enumerate(explanation["feature_names"]):
+        x_feat = feat_values[i, :, 0]
+        trend_fit = LinearRegression()
+        trend_line = trend_fit.fit(x_feat.reshape(-1, 1), explanation["signed_impact"])
+        x_trend = np.linspace(min(x_feat), max(x_feat), 101)
+        y_trend = trend_line.predict(x_trend[..., np.newaxis])
+
+        # Scatter plot
+        axs[i].scatter(x=x_feat, y=explanation["signed_impact"], color='blue')
+        # Line plot
+        axs[i].plot(x_trend, y_trend, color="green", label="correlation between forecast and observed feature")
+        # Reference line
+        current_value = explanation["current_feature_values"][i][0]
+        axs[i].axvline(x=current_value, color='firebrick', linestyle='--', label="current value")
+
+        axs[i].set_xlabel(feat)
+        axs[i].set_ylabel('Δ forecast')
+
+    # Display the legend on the first subplot
+    axs[0].legend()
+
+    fig.suptitle("Impact of Derived Variable On The Forecast", fontsize=16)
+    plt.tight_layout()
+    plt.subplots_adjust(top=0.95)
+    return fig
+
+def plot_ts(df, df_timestamps, df_timestamp_name, df_targets, df_description):
+    n_targets = len(df_targets)
+
+    fig, axs = plt.subplots(n_targets, 1, figsize=(10, 5 * n_targets))
+
+    # In case there's only one target, make sure axs is a list
+    if n_targets == 1:
+        axs = [axs]
+
+    for ax, target in zip(axs, df_targets):
+        ax.plot(df_timestamps, df[target], color="black", label=target)
+        ax.set_xlabel(df_timestamp_name)
+        ax.set_ylabel(target)
+        ax.set_title(f"[target] {target}")
+        ax.legend()
+
+    fig.suptitle(df_description)
+    plt.tight_layout()
+    plt.subplots_adjust(top=0.95)
+    return fig