SPI Subsets

Description of all pre-defined subsets in pyspi.

As standard, we include several SPI subsets in pyspi that can be used directly when instantiating a Calculator. Below we provide details about each subset including its intended purpose.

Subset	# of SPIs	Description
fabfour	4	Four basic pairwise measures that are commonly used in the literature.
sonnet	14	A minimally redundant set obtained by grouping SPIs according to their empirical behaviour on over 1000 MTS. Further details are provided in this paper. We also provide short names for each of the features in this subset.
fast	216	A subset of SPIs that are fastest to compute. Provides an ideal balance between comprehensiveness of pairwise measures and computational demand.
all	284	Default option. Contains the entire library of SPIs.

Subset Usage

To specify a subset, simply initialise the Calculator using the subset name as a parameter:

from pyspi.calculator import Calculator
import numpy as np 

dat = np.random.randn(2, 100)
calc = Calculator(dataset=dat, subset='fast')
calc.compute()