Mercurial > octave-nkf
diff scripts/statistics/tests/welch_test.m @ 3457:e031284eea27
[project @ 2000-01-19 08:49:56 by jwe]
author | jwe |
---|---|
date | Wed, 19 Jan 2000 08:50:14 +0000 |
parents | 434790acb067 |
children | 74c2fc84f0cf |
line wrap: on
line diff
--- a/scripts/statistics/tests/welch_test.m Wed Jan 19 06:59:23 2000 +0000 +++ b/scripts/statistics/tests/welch_test.m Wed Jan 19 08:50:14 2000 +0000 @@ -26,8 +26,9 @@ ## interest can be selected. If @var{alt} is @code{"!="} or ## @code{"<>"}, the null is tested against the two-sided alternative ## @code{mean (@var{x}) != @var{m}}. If @var{alt} is @code{">"}, the -## one-sided alternative mean(x) > m is considered, similarly for -## @code{"<"}. The default is the two-sided case. +## one-sided alternative mean(x) > @var{m} is considered. Similarly for +## @code{"<"}, the one-sided alternative mean(x) < @var{m} is +## considered. The default is the two-sided case. ## ## The p-value of the test is returned in @var{pval}. ## @@ -54,7 +55,7 @@ v_x = sumsq (x - mu_x) / (n_x * (n_x - 1)); v_y = sumsq (x - mu_y) / (n_y * (n_y - 1)); c = v_x / (v_x + v_y); - df = 1 / ( c^2 / (n_x - 1) + (1 - c)^2 / (n_y - 1)); + df = 1 / (c^2 / (n_x - 1) + (1 - c)^2 / (n_y - 1)); t = (mu_x - mu_y) / sqrt (v_x + v_y); cdf = t_cdf (t, df);