brede_loc_test_gaussian

(export/brede/brede_loc_test_gaussian.m)

Function Synopsis

R = brede_loc_test_gaussian(L, varargin)

Help text

 brede_loc_test_gaussian - Test for gaussianity

       R = brede_loc_test_gaussian(L, varargin)

       Input:    L    'Loc' structure or matrix

       Property: Output  [ {Pvalue} | statistic ]
                 Type    [ {MinAB} | MardiaA | MardiaB ]

       Output:   R    Result

       This function test for multivariate normality (Gaussianity) of
       a set of locations.

       The minimum P-value for Mardia's test on skewness and kurtosis
       are returned as the result. The p-value is based on
       asymptotics. 

       The P-value for Mardia's B (kurtosis test) is converted to a
       two-sided interval by:
                              
                pValue = 2*(0.5-abs(lta - 0.5))

       where lta is the lower tail area.

       Ref: Mardia, Multivariate Analysis (1979) p 21, 148-149
            Mardia, Biometrika (1970) 57(3):519+

       See also BREDE, BREDE_LOC, BREDE_LOC_TEST_EQUAL,
                BREDE_LOC_LOC2MAT. 

 $Id: brede_loc_test_gaussian.m,v 1.1 2003/08/14 14:49:01 fnielsen Exp $

Cross-Reference Information

brede_cdf_chi2 export/brede/brede_cdf_chi2.m
brede_cdf_gaussexport/brede/brede_cdf_gauss.m
brede_loc_loc2matexport/brede/brede_loc_loc2mat.m