fix: handle ValueError in fast mode compilation for anomaly evaluation

[justinuliu]

The following code will cause an exception.

import numpy as np
from sklearn.datasets import make_blobs
import matplotlib.pyplot as plt
from capymoa.stream import NumpyStream, Schema
from capymoa.instance import LabeledInstance, RegressionInstance

from capymoa.evaluation.visualization import plot_windowed_results
from capymoa.anomaly import HalfSpaceTrees, Autoencoder, OnlineIsolationForest
from capymoa.evaluation import prequential_evaluation_anomaly


# generate normal data points
n_samples = 10000
n_features = 2
n_clusters = 3
n_anomalies = 500

X, y = make_blobs(n_samples=n_samples, n_features=n_features, centers=n_clusters, random_state=42)

# generate anomalous data points
anomalies = np.random.uniform(low=-10, high=10, size=(n_anomalies, n_features))

# combine the normal data points with anomalies
X = np.vstack([X, anomalies])
y = np.hstack([y, [1]*n_anomalies])  # Label anomalies with 1
y[:n_samples] = 0  # Label normal points with 0


# create a NumpyStream from the dataset
feature_names = [f"feature_{i}" for i in range(n_features)]
target_name = "class"

stream = NumpyStream(X, y, dataset_name="AnomalyDetectionDataset",
                       feature_names=feature_names, target_name=target_name)
hst = HalfSpaceTrees(schema=stream.get_schema())

results_hst = prequential_evaluation_anomaly(stream=stream, learner=hst, window_size=100, store_predictions=True, store_y=True, optimise=False, max_instances=1000)

print(f"[HalfSpaceTrees] AUC: {results_hst.auc()} wallclock: {results_hst.wallclock()}")

The reason is that NumpyStream is incompatible with fast mode and raises a ValueError exception. We can handle this exception and continue using the default mode.

[tachyonicClock]

Thanks for identifying this. I changed NumpyStream.get_moa_stream() to return None if the feature is unsupported in #251 . _is_fast_mode_compilable already understands None and will correctly use python mode.