lyngby_ts_uttest - Student's t-test for unequal variances.
LYNGBY_TS_UTTEST(X1, X2) gives the probability that Student's t
calculated on data X1 and X2, sampled from distributions
with different variances, is higher than observed, i.e.
the "significance" level. This is used to test whether
two sample have significantly different means.
[P, T] = LYNGBY_TS_UTTEST(X1, X2) gives this probability P and the
value of Student's t in T. The smaller P is, the more
significant the difference between the means.
E.g. if P = 0.05 or 0.01, it is very likely that the
two sets are sampled from distributions with different
means.
This works if the samples are drawn from distributions with
DIFFERENT VARIANCE. Otherwise, use LYNGBY_TSTTEST.
See also: LYNGBY_TS_TTEST, LYNGBY_TS_PTTEST.
Date Author Version
1997-06-01 Cyril Goutte Initial Code
1997-06-15 Peter Toft Modified to fit lyngby + optimization