Thanks Jan, that seems to work to calculate the points in the circle.
Using GeodeticCalculator, the setDistance takes a 'distance' parameter. I
want to set this in KM, but am not sure if any conversion is necessary.
@param distance The orthodromic distance in the same units as the
[EMAIL PROTECTED] #getEllipsoid ellipsoid} axis.
Can anyone explain this to me? Or point me to any existing information.
Thanks,
Cliff
On Tue, Apr 22, 2008 at 5:02 AM, [EMAIL PROTECTED] <
[EMAIL PROTECTED]> wrote:
> Hello Cliff,
>
> In the case of a geodetic reference system (ie WGS84) if I'm not mistaken
> there is a utility class to do the apply the same solution:
> GeodeticCalculator. I never used it, but looking to the api it's what you
> need.
>
> 1) Instantiate this class with WGS84.
> 2) set the center point: setStartingGeographicPoint(double longitude,
> double latitude)
> 3) for each angle/azimuth:
> * setDirection(double azimuth, double distance)
> * getDestinationGeographicPoint()
>
> This means you'll calculate n points around the same point at the same
> distance at a different angle. The definition of a circle. The problem of
> the projection remains however. You'll get a perfect circle at the center of
> the projection but the more you'll go to the poles, the more "pear-shaped"
> your circle will be. On the poles you'll get very weird results. Especially
> for big circles. This is because on the cylindrical projection distances
> decrease at a rate of - roughly - cosine(latitude) while on the screen it's
> still the same dimension. The top line of the projection is in fact one
> single geographical point. All this explanation to tell this will only work
> properly in certain conditions. Otherwise you'll need to find another
> projection.
>
> If you need to have a perfect circle at all times, you'll need to inverse
> the problem: draw a circle on the screen in pixel coordinates and calculate
> for each point the geographical location - taking the projection and
> reference system into account. Unfortunately I don't know how you can do
> this in Geotools. Not yet.
>
> Jan
>
> >----- Oorspronkelijk bericht -----
> >Van: Cliff Darling [mailto:[EMAIL PROTECTED]
> >Verzonden: maandag, april 21, 2008 11:32 PM
> >Aan: 'Jan Goyvaerts'
> >CC: [email protected]
> >Onderwerp: Re: [Geotools-gt2-users] Create polygon/geometry for circle
> with given center and radius
> >
> >Thanks Jan.
> >
> >I am using a WGS84 projection (the default geographic projection). Is
> there
> >a way to modify the algorithm to work with a non-cartesian CRS? I am not
> >sure what would be involved.
> >
> >Would it suffice to convert to Cartesian, collect all of the points, and
> >then transform back to WGS84?
> >
> >Thanks,
> >
> >Cliff
> >
> >
> >
> >On Mon, Apr 21, 2008 at 10:07 PM, Jan Goyvaerts <[EMAIL PROTECTED]
> >
> >wrote:
> >
> >> Hi Cliff,
> >>
> >> No Java code at hand, but in pseudo code this should basically look
> like
> >> this:
> >>
> >> for angle = 0 to 359 step 1 do
> >> x = centerX + raduis * sine(angle)
> >> y = centerY + radius * cosine(angle)
> >> done
> >>
> >> (Decrease the step value of the for loop to make it look smoother.)
> >>
> >> This is assuming the coordinate reference system is carthesian. The
> >> projection you'll use for the map is of importance too. As it might not
> >> display the geometry as a round circle.
> >>
> >> Jan
> >>
> >> On Mon, 21 Apr 2008 20:16:16 +0100
> >> "Cliff Darling" <[EMAIL PROTECTED]> wrote:
> >>
> >> > Does anyone have a code snippet for creating a polygon that
> represents a
> >> > circle around a center point with a given radius?
> >> >
> >> > Thanks,
> >> >
> >> > Cliff
> >> >
> >>
> >>
> >> --
> >> Jan Goyvaerts <[EMAIL PROTECTED]>
> >>
> >>
> -------------------------------------------------------------------------
> >> This SF.net email is sponsored by the 2008 JavaOne(SM) Conference
> >> Don't miss this year's exciting event. There's still time to save $100.
> >> Use priority code J8TL2D2.
> >>
> >>
> http://ad.doubleclick.net/clk;198757673;13503038;p?http://java.sun.com/javaone
> >> _______________________________________________
> >> Geotools-gt2-users mailing list
> >> [email protected]
> >> https://lists.sourceforge.net/lists/listinfo/geotools-gt2-users
> >>
> >
>
>
>
-------------------------------------------------------------------------
This SF.net email is sponsored by the 2008 JavaOne(SM) Conference
Don't miss this year's exciting event. There's still time to save $100.
Use priority code J8TL2D2.
http://ad.doubleclick.net/clk;198757673;13503038;p?http://java.sun.com/javaone
_______________________________________________
Geotools-gt2-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/geotools-gt2-users
