D'Agostino-Pearson test for normality

Description

Performs the D'Agostino-Pearson goodness-of-fit test for the hypothesis of normality. The null hypothesis is that the sample comes from a normal distribution.

Hypothesis of Normality The hypothesis of normality refers to the null hypothesis that the data comes from a normal distribution. In the implementation, the statistic is computed from the sample passed to execute_statistic.

Test Statistic The statistic is based on the sum of squared transformed skewness and kurtosis statistics.

Usage

from pysatl_criterion.statistics.goodness_of_fit import (
    DAPNormalityGofStatistic,
)


test_statistic = DAPNormalityGofStatistic()
statistic_result = test_statistic.execute_statistic([-1.21, -0.83, -0.52, -0.31, -0.08, 0.14, 0.29, 0.47, 0.68, 0.91, 1.16, 1.43])
print(statistic_result)

Arguments

mean - reference normal mean. Default value is 0.

var - reference normal variance. Default value is 1.

rvs - array-like sample data passed to execute_statistic.

Details

The implementation evaluates the D'Agostino-Pearson statistic for the supplied observations. Large or small values should be interpreted according to the statistic alternative used by the class implementation.

References

The statistic follows the implementation in pysatl_criterion.statistics.goodness_of_fit.normal.

Author(s)

Alexey Mironov

Examples

from pysatl_criterion.statistics.goodness_of_fit import (
    DAPNormalityGofStatistic,
)


test_statistic = DAPNormalityGofStatistic()
statistic_result = test_statistic.execute_statistic([-1.21, -0.83, -0.52, -0.31, -0.08, 0.14, 0.29, 0.47, 0.68, 0.91, 1.16, 1.43])
print(statistic_result)