cell2net.preprocessing.collapse_consecutive_values#

cell2net.preprocessing.collapse_consecutive_values(X)#

Collapse consecutive identical values in an array.

This function identifies segments of consecutive identical values in an input array and returns their start indices, unique values, and lengths of each segment.

Parameters:

X (ndarray) – A 1D array of values (integers or floats) to process.

Return type:

tuple[ndarray, ndarray, ndarray]

Returns:

A tuple containing:

  • idx (numpy.ndarray): Start indices of each segment of consecutive identical values.

  • values (numpy.ndarray): The unique values corresponding to each segment.

  • lengths (numpy.ndarray): The lengths (number of repetitions) of each segment.

Notes

  • This function is optimized for performance using numba.prange for parallel processing of the input array.

  • The output idx array contains the indices where each segment starts in X.

  • The values array contains the unique values from the input array, and the lengths array contains the counts of consecutive occurrences of each value.

  • To reconstruct the original input, use values.repeat(lengths).

Examples

>>> import numpy as np
>>> import cell2net as cn
>>> X = np.array([1, 1, 2, 2, 2, 3, 1, 1])
>>> idx, values, lengths = cn.pp.collapse_consecutive_values(X)
>>> print(idx)
... [0 2 5 6]
>>> print(values)
... [1. 2. 3. 1.]
>>> print(lengths)
... [2 3 1 2]
>>> np.array_equal(X, values.repeat(lengths))
... True