No, the buffer does not have any "inner points" (in other words, the input polygon is not a hole in the buffer - the buffer is a larger polygon that completely covers the input).
On Mon, Oct 26, 2015 at 2:31 PM, Jeff Dege <[email protected]> wrote: > Can I assume that the inner points of the buffer are identical to the > outer points of the source polygon? > > > > > > *From:* Martin Davis [mailto:[email protected]] > *Sent:* Monday, October 26, 2015 4:27 PM > *To:* Jeff Dege > *Cc:* [email protected] > *Subject:* Re: [Jts-topo-suite-user] Rounding error gaps with buffers? > > > > 1. The returned value from the Geometry.buffer method contains only the > buffer polygon. > > > > 2. If you use a "large enough" buffer distance the buffer polygon should > properly contain the input polygon. And the "large enough" distance can be > quite small, since double precision is used during computation. > > > > If you want to try out how JTS responds to your inputs, I suggest > installing the product and using the JST TestBuilder to try things out. > > > > Martin > > > > On Mon, Oct 26, 2015 at 2:01 PM, Jeff Dege <[email protected]> wrote: > > We have a problem where we need to draw areas on a map that only slightly > overlap. This is used in a routing application, where events are assigned > to the assigned user for each area. What we want is to have the areas > overlap slightly, so as to avoid any possibility of there being a gap. > > > > For us to have a small area along the border between two areas in which > which user the event is assigned to is acceptable. For us to have even the > smallest area in which an event might fall between the cracks is not. Given > my understanding of floating point rounding errors, it’s not possible to > have absolute equality. If you’re going to avoid any cracks, you’re going > to have to have some overlap. > > > > What we’re envisioning is to have a newly drawn area clip against the > existing area, and then to expand the newly drawn area slightly. > > > > We’re considering the JTS Topology Suite. It looks as if its Buffer > functionality might do what we need. But before we install it and try to > work with it, we have a few questions: > > > > 1. When I create a buffer around a polygon or multipolygon, does > the geometry that it returns contain the buffer plus the polygon, or just > the buffer? > > 2. If the returned geometry contains just the buffer, will a union > of the buffer with the original polygon never contain holes along the > boundary, due to floating point rounding errors? > > > > > > >
------------------------------------------------------------------------------
_______________________________________________ Jts-topo-suite-user mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/jts-topo-suite-user
