seg1d.Segmenter.add_reference

Segmenter.add_reference(r, copy=True)

Appends a reference containing one or more features to the existing reference dataset. If the reference is not a dict, it will be converted to one. If this should be the only reference set, use clear_reference() before calling this method.

Parameters:

rdict or ndarray

Dictionary containing labeled features as keys and values as 1-D arrays (must be same size).

ndarray of dimension 1 will be used as a single feature for the reference.

ndarray of n-dimensions will use rows as unique features.

copybool, optional

If True, will make a deepcopy of the passed parameter (Default True).

See also

set_target

Set the target data

clear_reference

Clear the current reference data

Notes

This method allows features that are not in previous references to be added, and vice-versa. It will also allow different sizes of reference data to be added. This is done as you can explicitly declare which features to use when segmenting.

Examples

Add a reference with multiple features

>>> import seg1d
>>> import numpy as np
>>> 
>>> s = seg1d.Segmenter()
>>> r = np.linspace(0,1,6).reshape(2,3)
>>> s.add_reference( r )
>>> s.r
[{'0': array([0. , 0.2, 0.4]), '1': array([0.6, 0.8, 1. ])}]

Alternatively, each row of the array can be added as the same labeled feature for different references by calling this method in a loop. Notice this is now an array of dictionaries containing the same feature label.

>>> s = seg1d.Segmenter()
>>> r = np.linspace(0,1,6).reshape(2,3)
>>> for _r in r: s.add_reference(_r)
>>> s.r
[{'0': array([0. , 0.2, 0.4])}, {'0': array([0.6, 0.8, 1. ])}]