úÎB¶@™      experimentalDaniël de Kok <me@danieldk.eu> Safe-InferredBonferroni correction: ±* = ± / n, where ± is the intended F significance level over all tests, n the number of tests, and ±* the  corrected significance level. &Sidak correction: ±* = 1 - (1 - ±)^(1 / n), where ± is the intended F significance level over all tests, n the number of tests, and ±* the  corrected significance level. Significance level Number of comparisons !The corrected significance level Significance level Number of comparisons !The corrected significance level  experimentalDaniël de Kok <me@danieldk.eu>NoneBA test stastic calculates the difference between two samples. See   and  for examples. Significance. ,Data is compatible with the null hypothesis 'The null hypothesis should be rejected "The result of hypothesis testing.  Significance Test statistic for the samples Test statistics for the  randomized samples  Options for randomization tests Type of test ( or ) Test statistic %Number of shuffled samples to create (he p-value at which to test (e.g. 0.05) 0Computations with random numbers that can fail. 2Apply a pair-wise approximate randomization test. HIn pair-wise approximate randomization tests the data points at a given D index are swapped between samples with a probability of 0.5. Since @ swapping is pairwise, the samples should have the same length. )Apply an approximate randomization test. BIn approximate randomization tests, the values of two samples are B shuffled among those samples. A test statistic is calculated for F the original samples and the shuffled samples, to detect whether the . difference of the samples is extreme or not. Determine the significance. Wrap a p-value in a . Calculate a p-value HCount extreme test statistic values. If the test statistic value of the H original sample is in the right tail, we want to count values equal to I or larger than that value. If the value is in the left tail, we want to G count value smaller than or equal to that value. Since we do not know  the tail (yet), we count both. HNote: we can determine the tail by (1) averaging the test statistics of F the randomized samples, or (2) taking the smaller of the two counts. DGenerate a given number of pairwise shuffled samples, and calculate & the test statistic for each shuffle. JSince the data points at a given index are swapped (with a probability of 0 0.5), the samples should have the same length. DGenerate a given number of shuffled samples, and calculate the test  statistic for each shuffle. DThis function does not require the samples to have an equal length. "Pair-wise shuffle of two vectors. /Calculates the difference mean of two samples ( mean(s1 - s2) ). When the G two samples do not have an equal length, the trailing elements of the  longer vector are ignored. /Calculates the mean difference of two samples (mean(s1) - mean(s2)).  Calculate the mean of a sample. )Calculate the ratio of sample variances (var(s1) : var(s2)). !Subtract two vectors. "$Generate Int numbers within a range % Options for the test  First sample Second sample The test result Options for the test  First sample Second sample The test result Type of test ( or ) )The p-value at which to test (e.g. 0.05) Number of sample shuffles Extreme statistic counts The test result The p-value at which to test  The p-value The test result Number of extreme outcomes Number of shuffles  The p-value -Test statistic value of the original samples 1Test statistic values of the randomized samples. (Count of left- and right-tail extremes. Test statistic %Number of shuffled samples to create  First sample Second sample The statistics of the shuffles Test statistic %Number of shuffled samples to create  First sample Second sample The statistics of the shuffles #$%& !'("     #$%& !'(")        !"#$%&'()*+,approx-rand-test-0.2.0Statistics.Test.CorrectionStatistics.Test.ApproxRand bonferronisidak TestStatistic SignificanceNotSignificant Significant TestResulttrSignificancetrStattrRandomizedStats TestOptions toTestTypetoTestStatistic toIterationstoPValue RandWithErrorapproxRandPairTestapproxRandTestapproxRandPairStatsapproxRandStatsdifferenceMeanmeanDifference varianceRatiostatistics-0.10.4.1Statistics.Test.Types OneTailed TwoTailed significance significantpValue countExtremesshuffleVectorsPairwisemean subVector randomIntR randomVectorshuffleVectors shuffleVectorshuffleVector'subIIWaddIWI