Hi Leonard,
Leonard Mada schrieb:
Hi Regina,
What did you decide with the z-test?
I have not decided anything yet.
I lost slightly the track.
I recommended extending the t-test. And I still believe this is the
better solution.
The descriptions of the t-test in the ODF draft spec seems very faulty
to me. But the intention is to define it in that way in which it is used
in existing spreadsheets.
While for z-test, you may want to implement another parameter specifying
the number of tails:
"lesser", "greater" or "two-tailed"
That would be a change in the ODF draft spec. At the moment it is
two-tailed. But your proposal would solve the naming problem. It seems
that it is necessary to contact the ODF formula group. But not now, it
is after midnight here.
So, older spreadsheets could be imported using the new function by
setting the appropriate tails (I believe the tails was the problem of
the original function - but did not test this assertion).
There are more problems with the old implementation:
1. The function now has a real bug
(http://qa.openoffice.org/issues/show_bug.cgi?id=90759).
2. The function claims to calculate a one-tailed solution, but does it
in a very strange way.
Then, there
would be only a problem when saving a new document in an old format.
The way from new to old format would be no problem, if the new format
will include the old format. If there are some options which are only
available in the new format, that would not matter. That is the same
situation as in ADDRESS with the new format parameter, or will be in
BETADIST with the new density option.
Thanks for your help,
Regina
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
