OASIS Mailing List ArchivesView the OASIS mailing list archive below
or browse/search using MarkMail.

 


Help: OASIS Mailing Lists Help | MarkMail Help

office-formula message

[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]


Subject: GAMMADIST function's mathematical formula


Hi,

I just got a heads-up from an OOo code contributor working on functions
that our mathematical formula of GAMMADIST is different from what Calc
and Excel implemented.

Our definition says

GAMMADIST(x;α;λ;FALSE()) = λ^α / Γ(α) * x^(α-1) * e^(-λ*x)

Implementation in Calc and Excel is

GAMMADIST(x;a;b;FALSE()) = 1 / (b^a * Γ(a) ) * x^(a-1) * e^(-x/b)

See also the ECMA/MOOXML draft.

The difference seems to result from literature where both definitions
can be encountered, with λ=1/b. However, results are slightly different,
e.g.

Implementation
GAMMADIST(2;3;5;FALSE()) returns ≈0.010725

whereas our definition
GAMMADIST(2;3;5;FALSE()) would return ≈0.01134998

I propose to change our definition to reflect reality of common
implementation.

  Eike

-- 
Automatic string conversions considered dangerous. They are the GOTO statements
of spreadsheets.  --Robert Weir on the OpenDocument formula subcommittee's list.

PGP signature



[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]