The linear_segment API reference#
bcpseg.bcpseg#
- linear_segment.segment(values, labels=None, truncate=-100, cutoff=0.75, method='online_both', hazard=100, offset=10, shuffles=1000, p=0.05)[source]#
Implementation of a Bayesian Change Point Detection algorithm to segment values into Intervals
- Parameters:
- values : numpy.ndarray
Floats to segment
- labels : numpy.ndarray
Labels for breakup segmentation
- truncate : float
Tolerance during offline segmentation [default=-100]
- cutoff : float
Probability threshold for determining segment bounds [default=0.75]
- method : str
Method to use: offline, online, online_both [default:’online’]
- hazard : float
Expected typical segment length [default:100]
- offset : int
Number to skip before calculating probability [default:10]
- shuffles : int
Number of shuffles conduct
- p : double
Pvalue cutoff
- Return type:
LabeledIntervalArray
- Returns:
- segmentsLabeledIntervalArray
Segment intervals