Probabilistic CUSUM Detector Class

class source.detector.cusum.ProbCUSUM_Detector(warmup_period: int = 10, threshold_probability: float = 0.001)

Bases: Cusum

Probabilistic CUSUM Change Point Detector. A class to detect change points in sequential data using a probabilistic approach based on the CUSUM algorithm.

Parameters:

  • warmup_period (int) – The warmup period for the detector. Must be equal or greater than 10.

  • threshold_probability (float) – The threshold probability for detecting a change point. Must be between 0 and 1.

Initializes the Probabilistic CUSUM Detector with the specified parameters.

Parameters:

  • warmup_period (int) – The warmup period for the detector. Must be equal or greater than 10.

  • threshold_probability (float) – The threshold probability for detecting a change point. Must be between 0 and 1.

Raises:

ValueError – If warmup_period < 10 or threshold_probability is not between 0 and 1.

__init__(warmup_period: int = 10, threshold_probability: float = 0.001)

Initializes the Probabilistic CUSUM Detector with the specified parameters.

Parameters:

  • warmup_period (int) – The warmup period for the detector. Must be equal or greater than 10.

  • threshold_probability (float) – The threshold probability for detecting a change point. Must be between 0 and 1.

Raises:

ValueError – If warmup_period < 10 or threshold_probability is not between 0 and 1.

detection(observation: float)

Predicts the probability of a change point in the next observation.

Parameters:

observation (float) – The next data point to predict.

Returns:

  • probability (float) – The probability of a change point.

  • is_changepoint (bool) – Indicates if a change point is detected.

offline_detection(data: ndarray)

Detects change points in the given data in an offline manner.

Parameters:

data (numpy.ndarray) – Data points to be analyzed.

Returns:

results – A dictionary containing: - ‘probabilities’: numpy.ndarray of probabilities for each observation. - ‘is_drift’: list of booleans indicating detected change points. - ‘change_points’: numpy.ndarray of detected change point indices.

Return type:

dict

plot_change_points(data: ndarray, change_points: list, probabilities: list)

Plots data with detected change points and probabilities.

Parameters:

  • data (numpy.ndarray) – Original data points.

  • change_points (list) – List of detected change points.

  • probabilities (list) – List of probabilities associated with each data point.

Examples

Instance-based Detection

from source.detector.cusum import ProbCUSUM_Detector

detector = ProbCUSUM_Detector(warmup_period=10, threshold_probability=0.01)
data_stream = np.concatenate([np.random.normal(0, 1, 100),
                    np.random.normal(5, 1, 100)])
for data in data_stream:
    prob, is_change = detector.detection(data)
    print(f"Change Detected: {is_change} \n -Probability: {prob[0]}")

Batch-based Detection

from source.detector.cusum import ProbCUSUM_Detector

detector = ProbCUSUM_Detector(warmup_period=10, threshold_probability=0.01)
data = np.concatenate([np.random.normal(0, 1, 100),
                    np.random.normal(5, 1, 100)])
results = detector.offline_detection(data)
detector.plot_change_points(data,
                            results["change_points"],
                            results["probabilities"])

Plotting

Probabilistic CUSUM Example