[Proj] truble to translate coordinates from WGS84 to EPSG:31287

[Mikael Rittri]

The matrix multiplication, scalar multiplication and the 
addition takes place in 3D geocentric Cartesian space. 

Since you say your Z is zero, I think you may have 
forgotten the conversion from long/lat/height to X,Y,Z 
which are geocentric Cartesian coordinates, usually in 
meters. (That's the step 2 briefly mentioned on the
Wikipedia page on Helmert transformation.) The geocentric
Z is zero only on the equator. 

And the result of the Helmert transformation must
be converted back from X,Y,Z to long/lat/height.

Formulas can be found in section 2.2.1 of EPSG Guidance Note 7.2,
http://www.epsg.org/guides/G7-2.html 

Best regards,
Mikael Rittri
Carmenta AB
Sweden
http://www.carmenta.com

-----Original Message-----
From: proj-bounces at lists.maptools.org [mailto:proj-bounces at lists.maptools.org] On Behalf Of Markus Hetzmannseder
Sent: den 29 oktober 2010 01:39
To: PROJ.4 and general Projections Discussions
Subject: Re: [Proj] truble to translate coordinates from WGS84 to EPSG:31287

On Thu, 28 Oct 2010, Jean-Claude Repetto wrote:

> Le 28/10/2010 15:49, Markus Hetzmannseder a écrit :
>
> Applying a 7-parameters datum transformation is not that difficult ! 
> It requires a matrix multiplication, a scalar multiplication and an addition :
> http://en.wikipedia.org/wiki/Helmert_transformation

Well, the formulars there are looking not to bad, at to do some calcs on a excel sheet. The truble is, i still dont get the correct results for my few example parameters.

The formular i used for the X part was:

Xb = Cx + (1 + s / 1000000) * (Xa - Rz*PI/648000 * Ya)

The PI/648000 should do the calc from arcsecond to rad used at the formula. The Ry * Za part i have skipped, because i have no hight, so it should be save to set it to zero.

For the testing i want to translate my example point from my 1st post
here:

WGS84                    without geodesy date trans   with date trans
14.2833092 48.3098392    470439.28      490437.40     470507.17  490503.03

The formular for the X part of that point is:

-577.326 + (1 - 2.423 / 1000000) * (470439.28 - 5.297*3.1415/648000 * 490437.40)

Results to 469848.219810 but it should be 470507.17

So where could be the error in that?
Why have all of the transform parameters at the cs2cs command line a "+" 
and not a "-" like some at the wiki page of the Helmert transformation? But the error here has to be more then just the sign alone, because the Rz part of the formular alone would just produce an offset of about 12.5 but from Cx i get a translation of 577.


Mar *the clueless newb in date transformation* kus