statsmodels.sandbox.stats.multicomp.set_remove_subs statsmodels.sandbox.stats.multicomp.set_remove_subs(ssli) [source] remove sets that are subsets of another set from a list of tuples Parameters: ssli : list of tuples each tuple is considered as a set Returns: part : list of tuples new list with subset tuples removed, it is sorted by set-length of tuples. The list contains original tuples, duplicate elements are not removed. Examples >>> set_remove_subs([(0, 1), (1, 2), (1,

statsmodels.sandbox.stats.multicomp.set_partition statsmodels.sandbox.stats.multicomp.set_partition(ssli) [source] extract a partition from a list of tuples this should be correctly called select largest disjoint sets. Begun and Gabriel 1981 don?t seem to be bothered by sets of accepted hypothesis with joint elements, e.g. maximal_accepted_sets = { {1,2,3}, {2,3,4} } This creates a set partition from a list of sets given as tuples. It tries to find the partition with the largest sets. That i

statsmodels.sandbox.stats.multicomp.rejectionline statsmodels.sandbox.stats.multicomp.rejectionline(n, alpha=0.5) [source] reference line for rejection in multiple tests Not used anymore from: section 3.2, page 60

statsmodels.sandbox.stats.multicomp.rankdata statsmodels.sandbox.stats.multicomp.rankdata(x) [source] rankdata, equivalent to scipy.stats.rankdata just a different implementation, I have not yet compared speed

statsmodels.sandbox.stats.multicomp.randmvn statsmodels.sandbox.stats.multicomp.randmvn(rho, size=(1, 2), standardize=False) [source] create random draws from equi-correlated multivariate normal distribution Parameters: rho : float correlation coefficient size : tuple of int size is interpreted (nobs, nvars) where each row Returns: rvs : ndarray, (nobs, nvars) where each row is a independent random draw of nvars-dimensional correlated rvs

statsmodels.sandbox.stats.multicomp.qcrit statsmodels.sandbox.stats.multicomp.qcrit = '\n 2 3 4 5 6 7 8 9 10\n5 3.64 5.70 4.60 6.98 5.22 7.80 5.67 8.42 6.03 8.91 6.33 9.32 6.58 9.67 6.80 9.97 6.99 10.24\n6 3.46 5.24 4.34 6.33 4.90 7.03 5.30 7.56 5.63 7.97 5.90 8.32 6.12 8.61 6.32 8.87 6.49 9.10\n7 3.34 4.95 4.16 5.92 4.68 6.54 5.06 7.01 5.36 7.37 5.61 7.68 5.82 7.94 6.00 8.17 6.16 8.37\n8 3.26 4.75 4.04 5.64 4.53 6.20 4.89 6.62 5.17 6.96 5.40 7.24 5.60 7.47 5.77 7.68 5.92 7.86\n9 3.20 4.60 3.9

statsmodels.sandbox.stats.multicomp.multipletests statsmodels.sandbox.stats.multicomp.multipletests(pvals, alpha=0.05, method='hs', is_sorted=False, returnsorted=False) test results and p-value correction for multiple tests Parameters: pvals : array_like uncorrected p-values alpha : float FWER, family-wise error rate, e.g. 0.1 method : string Method used for testing and adjustment of pvalues. Can be either the full name or initial letters. Available methods are `bonferroni` : one-step

statsmodels.sandbox.stats.multicomp.MultiComparison class statsmodels.sandbox.stats.multicomp.MultiComparison(data, groups, group_order=None) [source] Tests for multiple comparisons Parameters: data : array independent data samples groups : array group labels corresponding to each data point group_order : list of strings, optional the desired order for the group mean results to be reported in. If not specified, results are reported in increasing order. If group_order does not contain a

statsmodels.sandbox.stats.multicomp.mcfdr statsmodels.sandbox.stats.multicomp.mcfdr(nrepl=100, nobs=50, ntests=10, ntrue=6, mu=0.5, alpha=0.05, rho=0.0) [source] MonteCarlo to test fdrcorrection

statsmodels.sandbox.stats.multicomp.maxzerodown statsmodels.sandbox.stats.multicomp.maxzerodown(x) [source] find all up zero crossings and return the index of the highest Not used anymore >>> np.random.seed(12345) >>> x = np.random.randn(8) >>> x array([-0.20470766, 0.47894334, -0.51943872, -0.5557303 , 1.96578057, 1.39340583, 0.09290788, 0.28174615]) >>> maxzero(x) (4, array([1, 4])) no up-zero-crossing at end >>> np.random.seed(0) &