gdalogr
We use ArcObjects pretty heavily and simplify polygons using something similar to the ITopologicalOperation::Simplify(void). No tolerance is passed in as the parameter (it uses the XY tolerance of the associated spatial reference).
Now, we're using OGR for some functionality and would like to simplify using this library. What's the equivalent? I'm aware of OGRGeometry::SimplifyPreserveTopology(double dTolerance) but I'm not sure where to get the tolerance from the in OGR library.
testareageom.GetGeometryRef(0).GetPoint(0)[0] will get you the X coord of the first point.
For completeness sake, if you were after the centroid of the polygon or the coordinates of all the points that comprise the polygon, you could use something like this:
centroid=testareageom.Centroid()
x = centroid.GetX()
print x
for ring in testareageom:
for point in xrange(ring.GetPointCount()):
x,y,z = ring.GetPoint(point)
print x
OpenJUMP Plus has a special tool for that purpose.
It works well if the source data is topologically good and really have shared vertices. OpenJUMP has also tools for improving topology like "Adjust polygon boundaries" that removes small gaps and overlaps and adds vertices to make the boundaries match.
Tools "Find coverage gaps" and "Find coverage overlaps" help to verify if the topology is OK.
A very reliable topology check is to create a linear graph from the polygons and then create a new polygon layer from the closed areas of the graph. The result should contain the same polygons as the source layer.
Best Answer
You must supply the tolerance, whether in Arc or OGR. It is the measure of your simplification operation and is just a decimal number in the the same units as the SRS of your data.
In Arc ESRI stores an XY tolerance as a separate value. This is proprietary to ESRI and not inherent in the SRS. Either you specify it or ESRI supplies a default - either way it is the same thing as you explicitly supplying a value to OGR. ESRI calculates the default but there is nothing sacred about the value. It's just a suggestion and may not be relevant to your particular data depending on its provenance, collection methods etc.