/montecarlino

Parellel implementation of bootstrap confidence intervals and permutation tests.

Primary LanguagePython

Montecarlino

Parellel implementation of bootstrap confidence intervals and permutation tests.

Installation

pip install montecarlino

Examples

Let's compute the 95% confidence interval for the median of a set of values:

from montecarlino import bootstrap_interval

x1 = np.array([0.80, 0.83, 1.89, 1.04, 1.45, 1.38, 1.91, 1.64, 0.73, 1.46])
bootstrap_interval(np.median, x1)

Now let's compare the medians of two samples, i.e. what's the p-value of the difference of the medians assuming there is no difference?

from montecarlino import grouped_permutation_test

def _median_difference(x, y):
    return np.median(x) - np.median(y)

x1 = np.array([0.80, 0.83, 1.89, 1.04, 1.45, 1.38, 1.91, 1.64, 0.73, 1.46])
x2 = np.array([1.15, 0.88, 0.90, 0.74, 1.21])
grouped_permutation_test(median_difference, [x1, x2])