Re: [office-comment] ODFF: list of suggestions

[Eike Rathke <erack@sun.com>]

--qftxBdZWiueWNAVY
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable

Hi David,

On Friday, 2008-05-02 09:10:00 +0100, David King wrote:

> B function
> * "this function returns the probability that the number of successful=20
> trials shall lie between S and S2." - should say clearly if inclusive or=
=20
> exclusive (inclusive I think).
> * Test cases:   " =3DB(7;0.5;3)   0.24343750" - wrong result - 0.2734375=
=20
> is correct, as returned by Calc, and checked from the actual formula by=
=20
> me.

Done for draft11.


> NEGBINOMDIST
> * The formula image could be improved - k is undefined, and unnecessary
> *  This really needs to say "exactly number_f failures"
> *  For consistency, could simply use eg "F" rather than "number_f" in=20
> parameters. Even better, use standard symbols x, r, p.

Done for draft11.


> GAUSS
> NORMSDIST (mentioned twice) should read LEGACY.NORMSDIST. Might be=20
> better to use NORMDIST for the definition instead.

Done for draft11.


> NORMSDIST
> " exactly NORMDIST(x)." - NORMDIST is not defined with only 1 parameter=
=20
> - should be "NORMDIST(x;0;1;TRUE())".

Done for draft11, in LEGACY.NORMSDIST of course.


> LOGNORMDIST
> *Syntax: "NORMDIST " should read "LOGNORMDIST"

Done for draft11.

> *Formulae: - I suspect that the integral may need to be from -infinity=20
> not 0 for the cumulative dist function (as with NORMDIST). I can't find=
=20
> this formula on the net - normally it seems to be quoted as=20
> 1/2+1/2*ERF((LN(x)-u)/(s*SQRT(2))) - which does seem to produce the same=
=20
> results, and may be preferable if it is standard practice. Or could be=20
> defined in terms of NORMDIST or NORMSDIST which is the simplest.

Put on the TODO list.


> FV
> The first link to PV is dead

Already done earlier.


> FDIST
> Is it worth mentioning that as well as Calc, Gnumeric and KSpread, Excel=
=20
> (Excel97 at least) also seems to calculate the right tail?

Done for draft11.


> PROB
> *This needs to specify if Data must be ordered or not (unordered?).
> *It would also be better to be clear about whether Data must be=20
> sequential (is 1, 2, 3, 5 (missing 4) OK?) and/or integer (is 1, 2.5, 3,=
=20
> 4 OK?).
> *It would also be better to be clear about whether Start and End need to=
=20
> exist within Data (no?) and also whether they can be higher/lower than=20
> the max/min values in Data (yes?)
> * It would also be better to be clear about whether Start can be higher=
=20
> than End. Although they're defined as lower/upper bounds, is 'Start>End'=
=20
> an error, or valid with a probability of 0 returned?
> * This needs to specify how the result is calculated if End is omitted=20
> and Start is not present in Data (Calc returns 0).
> * This needs to specify the type of error returned - eg #N/A if Data,=20
> Probability are different lengths. This is important because #N/A is a=20
> specific (testable) type of error.

Put on the TODO list.


> "Array Functions"
> This categorisation seems to me to be inconsistent. Functions which=20
> return arrays are:
> MINVERSE, MMULT, MUNIT, TRANSPOSE, FREQUENCY, GROWTH, LINEST, LOGEST,=20
> TREND
> Functions which have ForceArray parameters are:
> MDETERM, MINVERSE, MMULT, SUMPRODUCT, SUMX2MY2, SUMX2PY2, SUMXMY2,=20
> CORREL, COVAR, FORECAST, FTEST, INTERCEPT, MODE, PEARSON, PROB, RSQ,=20
> SLOPE, STEYX, TTEST (and LOOKUP sometimes).
> This doesn't relate to the current functions listed as array functions.=
=20
> We have for instance SUMPRODUCT listed as an array function (but it's=20
> really a maths function) but MODE listed in stats (OK) - but both force=
=20
> array_mode calculation. Surely either both should be 'array functions'=20
> or both where they belong.
> Perhaps only matrix functions should be in this category: MINVERSE,=20
> MMULT, MUNIT, TRANSPOSE, MDETERM. It could then be called 'matrix=20
> functions'.

Put on the TODO list.

  Eike

--=20
 OpenOffice.org / StarOffice Calc core developer and i18n transpositionizer.
 SunSign   0x87F8D412 : 2F58 5236 DB02 F335 8304  7D6C 65C9 F9B5 87F8 D412
 OpenOffice.org Engineering at Sun: http://blogs.sun.com/GullFOSS

--qftxBdZWiueWNAVY
Content-Type: application/pgp-signature
Content-Disposition: inline

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.6 (GNU/Linux)

iD8DBQFJTq0BZcn5tYf41BIRAs+IAJ4oXSVz7ETsGopYz4MYePhP7VwPKgCeOZYB
E63/2MVANiXNldG17Kazr0g=
=9jzM
-----END PGP SIGNATURE-----

--qftxBdZWiueWNAVY--