Statistics::Table::F does not seem to be a two tailed Fisher Exact test (but maybe I am missing something). I had searched cpan documentation for Fisher and came up empty handed. By the way there is a perl version of a one tailed Fisher exact test here.
| [reply] |
AFAIK, the ANOVA is a Fisher Exact test,
two tailed. Aha, that test is a variation on the
chi square. OK, I completely misunderstood your question.
I'm only used to ANOVA kind of statistics ....
(About that Statistics::Table::F module:) I must admit the docs of that module are very minimal. Moreover,
I tried
to install it, but that doesn't work. I'm sorry that I've given
you a pointer that I never used myself.
The BSP package seems to have a really specialized use of
bigrams, I don't whether it can be used as-is for your
application. (update) After paging through it, I still can't easily
see where the bigram stops and the algoritm comes in.
To make it up to you, I will program your F-test.
Jeroen
"We are not alone"(FZ)
Update: There is javascript code to be
found here.
...time passes and I have parsed that javascript into perl:
Fisher's Exact test (probability of 2x2 tables)
| [reply] |
I believe the one tailed test used in the bigram analysis is in LeftFisher.pm
| [reply] |
