[Geotools-gt2-users] setting coordinate system of renderer

[Combe, C. <C....@na...>]

Hi,
Martin - thanks for your help, as you pointed out, I'm experiencing problems
caused by the renderer using a geographic (lat/long) coordinate system when
the data's based on a projected (cartesian) coordinate system. I've
attatched the code i'm using, some sample gml and the output from 2 attempts
to get it to work. 

My data uses the British National Grid so I copied code from
CreationTest.java (in the cts-coordtrans module) which sets up an
appropriate cs. Creating the cs appears to work (it produces the same output
as in the test), i then pass the new ProjectedCoordinateSystem object into
the constructor for StyledMapPane, which should create a map pane using that
cs. But when the renderer comes to display the map I get an error from
org.geotools.cs.Ellipsoid.orthodromicDistance Ellipsoid.java:382) (see
output1.txt), as Martin said:

>The fact that Ellipsoid.orthodromicDistance(...) 
>were invoked is an indication that the renderer is assuming a 
>non-cartesian CS. 

So, my nice new coordinate system isn't getting used. :( It looked to me as
if setCoordinateSystem(CoordinateSystem cs) in the Renderer class wasn't
really doing anything because the context hadn't been set and so no layers
needed to be changed - but i'm not sure about this.

As an alternative, i tried using the default constructor for StyledMapPane,
setting the context to the one which contains my 
data and then changing the cs of the renderer using:

   mapPane.getRenderer().setCoordinateSystem(projectCS);  

but the transformation from the default cs to the projected cs fails, giving
the message "Bursa Wolf parameters required", as 
changing the cs failed i then get the same error from the
orthodromicDistance method as before - see output2.txt. I haven't 
been able to get very far working out where these parameters should come
from or how to fix the problem - any help much appreciated,
cheers,
colin