statsmodels.sandbox.stats.runs.median_test_ksample statsmodels.sandbox.stats.runs.median_test_ksample(x, groups) [source] chisquare test for equality of median/location This tests whether all groups have the same fraction of observations above the median. Parameters: x : array_like data values stacked for all groups groups : array_like group labels or indicator Returns: stat : float test statistic pvalue : float pvalue from the chisquare distribution others ???? : currently some

statsmodels.sandbox.stats.runs.mcnemar statsmodels.sandbox.stats.runs.mcnemar(x, y=None, exact=True, correction=True) [source] McNemar test Parameters: x, y : array_like two paired data samples. If y is None, then x can be a 2 by 2 contingency table. x and y can have more than one dimension, then the results are calculated under the assumption that axis zero contains the observation for the samples. exact : bool If exact is true, then the binomial distribution will be used. If exact is f

statsmodels.sandbox.stats.runs.cochrans_q statsmodels.sandbox.stats.runs.cochrans_q(x) [source] Cochran?s Q test for identical effect of k treatments Cochran?s Q is a k-sample extension of the McNemar test. If there are only two treatments, then Cochran?s Q test and McNemar test are equivalent. Test that the probability of success is the same for each treatment. The alternative is that at least two treatments have a different probability of success. Parameters: x : array_like, 2d (N,k) dat

statsmodels.sandbox.stats.multicomp.varcorrection_unequal statsmodels.sandbox.stats.multicomp.varcorrection_unequal(var_all, nobs_all, df_all) [source] return joint variance from samples with unequal variances and unequal sample sizes something is wrong Parameters: var_all : array_like The variance for each sample nobs_all : array_like The number of observations for each sample df_all : array_like degrees of freedom for each sample Returns: varjoint : float joint variance. dfjoint

statsmodels.sandbox.stats.multicomp.varcorrection_unbalanced statsmodels.sandbox.stats.multicomp.varcorrection_unbalanced(nobs_all, srange=False) [source] correction factor for variance with unequal sample sizes this is just a harmonic mean Parameters: nobs_all : array_like The number of observations for each sample srange : bool if true, then the correction is divided by the number of samples for the variance of the studentized range statistic Returns: correction : float Correction

statsmodels.sandbox.stats.multicomp.varcorrection_pairs_unequal statsmodels.sandbox.stats.multicomp.varcorrection_pairs_unequal(var_all, nobs_all, df_all) [source] return joint variance from samples with unequal variances and unequal sample sizes for all pairs something is wrong Parameters: var_all : array_like The variance for each sample nobs_all : array_like The number of observations for each sample df_all : array_like degrees of freedom for each sample Returns: varjoint : array

statsmodels.sandbox.stats.multicomp.varcorrection_pairs_unbalanced statsmodels.sandbox.stats.multicomp.varcorrection_pairs_unbalanced(nobs_all, srange=False) [source] correction factor for variance with unequal sample sizes for all pairs this is just a harmonic mean Parameters: nobs_all : array_like The number of observations for each sample srange : bool if true, then the correction is divided by 2 for the variance of the studentized range statistic Returns: correction : array Corre

statsmodels.sandbox.stats.multicomp.TukeyHSDResults class statsmodels.sandbox.stats.multicomp.TukeyHSDResults(mc_object, results_table, q_crit, reject=None, meandiffs=None, std_pairs=None, confint=None, df_total=None, reject2=None, variance=None) [source] Results from Tukey HSD test, with additional plot methods Can also compute and plot additional post-hoc evaluations using this results class. Notes halfwidths is only available after call to plot_simultaneous. Other attributes contain infor

statsmodels.sandbox.stats.multicomp.tiecorrect statsmodels.sandbox.stats.multicomp.tiecorrect(xranks) [source] should be equivalent of scipy.stats.tiecorrect

statsmodels.sandbox.stats.multicomp.StepDown class statsmodels.sandbox.stats.multicomp.StepDown(vals, nobs_all, var_all, df=None) [source] a class for step down methods This is currently for simple tree subset descend, similar to homogeneous_subsets, but checks all leave-one-out subsets instead of assuming an ordered set. Comment in SAS manual: SAS only uses interval subsets of the sorted list, which is sufficient for range tests (maybe also equal variance and balanced sample sizes are requi