metrics

Metrics

Regression Metrics

For deterministic forecasts (point predictions with num_samples == 1), probabilistic forecasts (num_samples > 1), and quantile forecasts. For probabilistic and quantile forecasts, use parameter q to define the quantile(s) to compute the deterministic metrics on:

• Aggregated over time: Absolute metrics: - MERR: Mean Error - MAE: Mean Absolute Error - MSE: Mean Squared Error - RMSE: Root Mean Squared Error - RMSLE: Root Mean Squared Log Error

Relative metrics:
    - :func:`MASE <anomx.metrics.metrics.mase>`: Mean Absolute Scaled Error
    - :func:`MSSE <anomx.metrics.metrics.msse>`: Mean Squared Scaled Error
    - :func:`RMSSE <anomx.metrics.metrics.rmsse>`: Root Mean Squared Scaled Error
    - :func:`MAPE <anomx.metrics.metrics.mape>`: Mean Absolute Percentage Error
    - :func:`wMAPE <anomx.metrics.metrics.wmape>`: weighted Mean Absolute Percentage Error
    - :func:`sMAPE <anomx.metrics.metrics.smape>`: symmetric Mean Absolute Percentage Error
    - :func:`OPE <anomx.metrics.metrics.ope>`: Overall Percentage Error
    - :func:`MARRE <anomx.metrics.metrics.marre>`: Mean Absolute Ranged Relative Error

Other metrics:
    - :func:`R2 <anomx.metrics.metrics.r2_score>`: Coefficient of Determination
    - :func:`CV <anomx.metrics.metrics.coefficient_of_variation>`: Coefficient of Variation
    - :func:`AUTC <anomx.metrics.metrics.autc>`: Area Under Tolerance Curve

• Per time step: Absolute metrics: - ERR: Error - AE: Absolute Error - SE: Squared Error - SLE: Squared Log Error

Relative metrics:
    - :func:`ASE <anomx.metrics.metrics.ase>`: Absolute Scaled Error
    - :func:`SSE <anomx.metrics.metrics.sse>`: Squared Scaled Error
    - :func:`APE <anomx.metrics.metrics.ape>`: Absolute Percentage Error
    - :func:`sAPE <anomx.metrics.metrics.sape>`: symmetric Absolute Percentage Error
    - :func:`ARRE <anomx.metrics.metrics.arre>`: Absolute Ranged Relative Error

For probabilistic forecasts (storchastic predictions with num_samples >> 1) and quantile forecasts:

• Aggregated over time: Probabilistic metrics: - MCRPS: Mean Continuous Ranked Probability Score

Quantile metrics:
    - :func:`MQL <anomx.metrics.metrics.mql>`: Mean Quantile Loss
    - :func:`QR <anomx.metrics.metrics.qr>`: Quantile Risk

Quantile interval metrics:
    - :func:`MIW <anomx.metrics.metrics.miw>`: Mean Interval Width
    - :func:`MWS <anomx.metrics.metrics.miws>`: Mean Interval Winkler Score
    - :func:`MIC <anomx.metrics.metrics.mic>`: Mean Interval Coverage
    - :func:`MINCS_QR <anomx.metrics.metrics.mincs_qr>`: Mean Interval Non-Conformity Score for Quantile Regression

• Per time step: Probabilistic metrics: - CRPS: Continuous Ranked Probability Score

Quantile metrics:
    - :func:`QL <anomx.metrics.metrics.ql>`: Quantile Loss

Quantile interval metrics:
    - :func:`IW <anomx.metrics.metrics.iw>`: Interval Width
    - :func:`WS <anomx.metrics.metrics.iws>`: Interval Winkler Score
    - :func:`IC <anomx.metrics.metrics.ic>`: Interval Coverage
    - :func:`INCS_QR <anomx.metrics.metrics.incs_qr>`: Interval Non-Conformity Score for Quantile Regression

Classification Metrics

For deterministic forecasts (point predictions with num_samples == 1), probabilistic forecasts (num_samples > 1), and class label probability forecasts. Most metrics allow to extract the scores for specific labels with parameter labels.

• Aggregated over time: - Accuracy: Accuracy Score - Precision: Precision Score - Recall: Recall Score - F1: F1 Score - CM: Confusion Matrix

Dynamic Time Warping Metrics

For Dynamic Time Warping (DTW) (aggregated over time):

• DTW: Dynamic Time Warping Metric