[OSRS-PROJ] Re: Krovak / PROJ4 (+axis option?)

Radim Blazek blazek at itc.it
Wed Apr 30 08:16:42 PDT 2003


On Thursday 03 April 2003 07:38 pm, you wrote:
> Rather than have a +czech switch specifically for the Krovak projections,
> and instead of altering the sign conventions from the "usual", I would like
> to introduce a generic mechanism at the pj_transform() level for alternate
> axis orientations.
>
> Something along the line of a +axis switch, where "+axis=en" means that
> x axis = 'e'ast, and y axis = 'n'orth.  Then a coordinate system string
> could specify alternate axis orientations as +axis=sw indicating that
> X is a "southing" (or negative northing) and Y is a westing (or negative
> easting).  I think this would be the right organization for Krovak right?

Just a note, I think that 'top' usually is not precisely 'north' and
sometimes may be very different azimuth, so it may be confusing.
I would prefer 'left', 'right', 'top', 'bottom'.

> I have no schedule for implementing generic axis support or migrating to
> use Geralds Krovak (or simplified libproj).  I will not have time to even
> look into this for several weeks.  Thus, for your own purposes you should
> hack your local GRASS built in a manner suitable to your needs in the
> meantime.

Copy of PROJ.4 was removed from GRASS (5.1) and added as a new dependency.
That means that I have to hack my local PROJ.4. I can do that, but other users?

So again I would prefer if you could apply my patch for PJ_krovak.c
for now. PJ_kocc.c and +axis is good long term solution.

Gerald I. Evenden"  wrote:
> I'm probably missing something but if you want positive x-y in
> the first quadrant why not specify a false easting/northing to
> push everthing up and right. No 'czech' option though.

I don't want positive x-y in the first quadrant, I want negative 
x-y in the fourth quadrant. 
I thing that PJ_kocc.c returns what I need (I did not manage to compile
it in PROJ.4 or GRASS). 
I was talking about PJ_krovak.c, currently present in PROJ.4 and 
it returns positive values (but not switched axis).
AFAIK +x_0/+y_0 cannot switch sign/quadrant. 

> The mathematical origin is at the shifted pole, up near Helsinki
> and the area of interest is south and west of there, thus the
> negative values in the natural solution. 

In PJ_kocc.c, not in PJ_krovak.c.

Radim

----------------------------------------
PROJ.4 Discussion List
See http://www.remotesensing.org/proj for subscription, unsubscription
and other information.



More information about the Proj mailing list