[R] problem with pip2d() from ptinpoly

[Waichler, Scott R]

Thank you for catching that, Boris.  I'm surprised, given that there is no mention of sense of direction in the package documentation.

Scott Waichler

> -----Original Message-----
> From: Boris Steipe [mailto:boris.steipe at utoronto.ca]
> Sent: Friday, April 18, 2014 3:49 PM
> To: Waichler, Scott R; r-help at r-project.org
> Subject: Re: [R] problem with pip2d() from ptinpoly
> 
> Apparently it matters whether your polygon is defined clockwise or
> counterclockwise.
> 
> A point outside your triangle is recognized ...
> > q2 <- matrix(c(594893.0,115435.0), ncol=2, byrow=T) pip2d(Vertices =
> > verts, Queries = q2)
> [1] 1
> 
> ... and defining the triangle in counterclockwise sense gives the expected
> behaviour.
> > v2 <- matrix(c(594891,115309,594891,117201,594444,117201), ncol=2,
> > byrow=T) pip2d(Vertices = v2, Queries = query)
> [1] 1
> 
> Cheers,
> B.
>
> On 2014-04-18, at 6:00 PM, Waichler, Scott R wrote:
> 
> > Hi,
> >
> > pip2d() doesn't seem to work correctly for me.  I have a plot of a
> triangle that a query point fits inside, but the point is defined as
> outside the polygon by pip2d.
> >
> > library(ptinpoly)
> > verts <- matrix(c(594891,115309,594444,117201,594891,117201), ncol=2,
> > byrow=T) query <- matrix(c(594885.0,115435.0), ncol=2, byrow=T)
> > pip2d(Vertices = verts, Queries = query)  # result = -1 # contrary to
> > -1 output of pip2d, plot shows point lies within triangle
> > plot(c(594400, 595000), c(115000, 117500), type="n") polygon(verts,
> > border="red") points(x=query[,1], y=query[,2], col="blue")
> >
> > Scott Waichler
> > Pacific Northwest National Laboratory
> > Richland, WA, USA