Thread Subject: POLY2CW ??? what happened to it?

I'm having trouble using
inpolygon

and I need poly2cw to sort it out.

Where is it?

"Paul " <par@ceri.memphis.edu> wrote in message
news:fh8m9n$hcb$1@fred.mathworks.com...
> I'm having trouble using
> inpolygon
>
> and I need poly2cw to sort it out.
>
> Where is it?

POLY2CW is a function in Mapping Toolbox, but INPOLYGON doesn't call POLY2CW
(at least not in R2007b, and not in previous versions as far as I remember).
What version of MATLAB are you using, what trouble are you experiencing with
INPOLYGON, and why do think you need POLY2CW to sort it out?

--
Steve Lord
slord@mathworks.com

"Steven Lord" <slord@mathworks.com> wrote in message
<fh8n8i$s4v$1@fred.mathworks.com>...
>
> "Paul " <par@ceri.memphis.edu> wrote in message
> news:fh8m9n$hcb$1@fred.mathworks.com...
> > I'm having trouble using
> > inpolygon
> >
> > and I need poly2cw to sort it out.
> >
> > Where is it?
>
> POLY2CW is a function in Mapping Toolbox, but INPOLYGON
doesn't call POLY2CW
> (at least not in R2007b, and not in previous versions as
far as I remember).
> What version of MATLAB are you using, what trouble are you
experiencing with
> INPOLYGON, and why do think you need POLY2CW to sort it out?
>
> --
> Steve Lord
> slord@mathworks.com
>
>

clear
close('all')

x1 = [1/4 1/2];
y1 = [1/2 3/4];
x2 = [1/2 3/4];
y2 = [1/4 1/2];

xv = [x1,x2];
yv = [y1,y2];




x = rand(1000,1);
y = rand(1000,1);

plot(xv,yv,'ro')

hold on
plot(x,y,'g.')

use = inpolygon(x,y,xv,yv);
plot(x(use),y(use),'r.')

%%%%%%%%%%%%%%%%%%%%%

you can see that when I merge the two segments of vertices,
then the area I want to isolate is not correct. This is
basically what my problem is since I have two track of
particles and I want to isolate the region between them as
inside a polygon

Paul wrote:
> [snip]
>
> clear
> close('all')
>
> x1 = [1/4 1/2];
> y1 = [1/2 3/4];
> x2 = [1/2 3/4];
> y2 = [1/4 1/2];
>
> xv = [x1,x2];
> yv = [y1,y2];
>
>
>
>
> x = rand(1000,1);
> y = rand(1000,1);
>
> plot(xv,yv,'ro')
>
> hold on
> plot(x,y,'g.')
>
> use = inpolygon(x,y,xv,yv);
> plot(x(use),y(use),'r.')
>
> %%%%%%%%%%%%%%%%%%%%%
>
> you can see that when I merge the two segments of vertices,
> then the area I want to isolate is not correct.
 > [snip]

It is correct. If you plot the polygon with lines between the vertices,
like this:

plot(xv,yv,'r')

you'll see that your polygon is self-intersecting in a kind of hourglass
shape. That explains your result.

--
Steve Eddins
http://blogs.mathworks.com/steve

Steve Eddins <Steve.Eddins@mathworks.com> wrote in message
<fh9mea$p6r$1@fred.mathworks.com>...
> Paul wrote:
> > [snip]
> >
> > clear
> > close('all')
> >
> > x1 = [1/4 1/2];
> > y1 = [1/2 3/4];
> > x2 = [1/2 3/4];
> > y2 = [1/4 1/2];
> >
> > xv = [x1,x2];
> > yv = [y1,y2];
> >
> >
> >
> >
> > x = rand(1000,1);
> > y = rand(1000,1);
> >
> > plot(xv,yv,'ro')
> >
> > hold on
> > plot(x,y,'g.')
> >
> > use = inpolygon(x,y,xv,yv);
> > plot(x(use),y(use),'r.')
> >
> > %%%%%%%%%%%%%%%%%%%%%
> >
> > you can see that when I merge the two segments of vertices,
> > then the area I want to isolate is not correct.
> > [snip]
>
> It is correct. If you plot the polygon with lines between
the vertices,
> like this:
>
> plot(xv,yv,'r')
>
> you'll see that your polygon is self-intersecting in a
kind of hourglass
> shape. That explains your result.
>
> --
> Steve Eddins
> http://blogs.mathworks.com/steve

This is EXACTLY why I was asking about POLY2CW routine. I
have to make the vertices of the polygon ordered clockwise.

Paul wrote:

>> It is correct. If you plot the polygon with lines between
> the vertices,
>> like this:
>>
>> plot(xv,yv,'r')
>>
>> you'll see that your polygon is self-intersecting in a
> kind of hourglass
>> shape. That explains your result.
>>
>> --
>> Steve Eddins
>> http://blogs.mathworks.com/steve
>
> This is EXACTLY why I was asking about POLY2CW routine. I
> have to make the vertices of the polygon ordered clockwise.
>

I'm not sure that POLY2CW, which is in the Mapping Toolbox as Steve Lord
said earlier, will help you. POLY2CW does not reorder the vertices of a
self-intersecting polygon to make it non self-intersecting.

--
Steve Eddins
http://blogs.mathworks.com/steve