Welcome to staticrab’s documentation!¶
Staticrab is a python package aimed at fast(ish) calculation of correlation coefficients in python.
correlation.py¶
The module for fast correlation computation.
-
staticrab.correlation.
chatterjee
(x: numpy.ndarray, y: numpy.ndarray) → float[source]¶ Computes the Chatterjee’s correlation measure.
The function computes the Chatterjee’s correlation measure. The algorithm is described at arXiv:1909.10140. It computes a deterministic version of the coefficients where the ties are not broken at random. To have the random breakage, randomly shuffle the input arrays before using this function.
- Parameters
x – array of float64
y – array of float64, cannot be constant
- Returns
Chatterjee’s correlation coefficient for x, y
- Return type
float
Examples
>>> a = np.array(range(5), dtype=np.float64) >>> chatterjee(a, a) 0.5 >>> a = np.array(range(2000), dtype=np.float64) >>> chatterjee(a, a) 0.9985007496251874
Notes
There appeared another python implementation at https://github.com/czbiohub/xicor but the computed valus are different.