Latest Contributions by DanPatterson_Retired

Forum a) Draggers & Droppers (The Dark Arts of Model Builder) Forum b) White space matters (Zen and the Art of ArcPy) Forum C) Far too much overhead (ArcObjects) And a amplfication to b)  syntax matters Splitting is not an option, the user needs to be "trained" to formulate their questions appropriately.. with tags or whatever. ... View more

could it have anything to do with the fact that the 2.6 version of Python is installed in a path such as c:\Python26\ArcGIS10.0....... and not what you would expect from a straight Python install, perhaps SciPy expects a normal install path ... View more

It isn't going to fit on a page anymore but David Wynne posted a useful module that will allow you to browse what is there without having to use Python's help(module) or dir(module) options http://forums.arcgis.com/threads/10395-Geoprocessing-modules-tools-functions-reference-listing perhaps a small volume will be needed.  The help files still contain the most information ... View more

Nice!...never used it, but it is better than >>> import arcpy >>> help(arcpy) >>> dir(arcpy) ... View more

I would concentrate on the online help as a start http://help.arcgis.com/en/arcgisdesktop/10.0/help/index.html?TopicName=welcome#/What_is_geoprocessing/002s00000001000000/ since it is updated as changes occur ... View more

might be interesting http://blogs.esri.com/Support/blogs/mappingcenter/archive/2010/01/21/Introducing-the-ArcGIS-Bump-Map-Tools.aspx ... View more

you can substitute your case for the one shown here http://forums.arcgis.com/threads/3062-multiply-two-rasters-but-ignore-no-data-values-from-first-raster ... View more

in 9.3 http://webhelp.esri.com/arcgisdesktop/9.3/index.cfm?id=6299&pid=6281&topicname=Min in 10 http://help.arcgis.com/en/arcgisdesktop/10.0/help/index.html#//009z0000007z000000.htm if this is what you mean ... View more

See Bill Huber's August 13 submission at http://forums.arcgis.com/threads/1887-MVP-Program-for-Discussion-Forums for cross-link ideas ... View more

No... Geoprocessing (gp and its incarnation within ArcMap) is the focus. Modelbuilder and Python are the means to the end...Modelbuilder can be used in itself or, when exported, to see how Python programming can be used to exploit the gp. One could argue that many field calculation operations involving shape geometry be included here, but they tend to go to the ArcMap General category. One could also argue that gp-arcobjects "stuff" be included here, however, I would keep the gp-MB-py grouping together. ... View more

For polygons, the spatial reference (SR) needs to be explicit.  The original GeometryErrorsDemo.py is shown below with an example SR and its new output for a triangle, square and rectangle ''' GeometryErrorsDemo.py Demonstrates calculating various properties for polygons ''' import arcpy pnt = arcpy.Point() triangle = [[0,0],[0,1],[1,0]] square = [[0,0],[0,1],[1,1],[1,0]] rectangle = [[0,0],[0,1],[2,1],[2,0]] polygons = [triangle, square, rectangle] labels = ["Triangle", "Square", "Rectangle"] array = arcpy.Array() polys = [] for i in range(len(polygons)):   a_poly = polygons   print "\n", labels," Coordinates: ", a_poly   for pair in a_poly:     pnt.X = float(pair[0])     pnt.Y = float(pair[1])     array.add(pnt)     print " X %20.16f  Y %20.16f" % (pnt.X, pnt.Y)   array.add(array.getObject(0))   #poly = arcpy.Polygon(array)   #low precision   #   SR = arcpy.SpatialReference("c:/temp/NAD 1983 UTM Zone 18N.prj")   #   poly = arcpy.Polygon(array, SR)   print "Polygon properties:"   print " area %20.16f\n length %s" % (poly.area, poly.length)   print " centroid %s\n true centroid %s " % (poly.centroid, poly.trueCentroid)   print " first point %s\n last point %s " % (poly.firstPoint, poly.lastPoint)   print " extent %s " % (poly.extent)   print " hull %s\n " % (poly.hullRectangle)   array.removeAll()   polys.append(poly) Results below Triangle  Coordinates:  [[0, 0], [0, 1], [1, 0]] X   0.0000000000000000  Y   0.0000000000000000 X   0.0000000000000000  Y   1.0000000000000000 X   1.0000000000000000  Y   0.0000000000000000 Polygon properties: area   0.5000000000000000 length 3.41421356237 centroid 0.333333333333333 0.333333333333333 NaN NaN true centroid 0.333333333333333 0.333333333333333 NaN NaN first point 0 0 NaN NaN last point 0 0 NaN NaN extent 0 0 1 1 NaN NaN NaN NaN hull 1 2.22044604925031E-16 0.5 -0.5 -0.5 0.5 0 1 Square  Coordinates:  [[0, 0], [0, 1], [1, 1], [1, 0]] X   0.0000000000000000  Y   0.0000000000000000 X   0.0000000000000000  Y   1.0000000000000000 X   1.0000000000000000  Y   1.0000000000000000 X   1.0000000000000000  Y   0.0000000000000000 Polygon properties: area   1.0000000000000000 length 4.0 centroid 0.5 0.5 NaN NaN true centroid 0.5 0.5 NaN NaN first point 0 0 NaN NaN last point 0 0 NaN NaN extent 0 0 1 1 NaN NaN NaN NaN hull 6.12323399573677E-17 1 1 1 1 0 0 0 Rectangle  Coordinates:  [[0, 0], [0, 1], [2, 1], [2, 0]] X   0.0000000000000000  Y   0.0000000000000000 X   0.0000000000000000  Y   1.0000000000000000 X   2.0000000000000000  Y   1.0000000000000000 X   2.0000000000000000  Y   0.0000000000000000 Polygon properties: area   2.0000000000000000 length 6.0 centroid 1 0.5 NaN NaN true centroid 1 0.5 NaN NaN first point 0 0 NaN NaN last point 0 0 NaN NaN extent 0 0 2 1 NaN NaN NaN NaN hull 6.12323399573677E-17 1 2 1 2 -2.22044604925031E-16 0 0 ... View more

David One would expect high precision as the default... perhaps next time, in the interim poly = arcpy.Polyline(array,arcpy.SpatialReference()) seemed to work for polylines and I presume polygons etc... perhaps, documentation on, arcpy.SpatialReference would be useful or go the high route by default.  If one isn't creating any output geometry (next lesson) the SR is not necessary.  A fix would be nice...thanks for getting on this. ... View more

Dan Requires ArcInfo license...nice if you have it http://help.arcgis.com/en/arcgisdesktop/10.0/help/index.html#//000800000006000000.htm ... View more

Contacted by ESRI (esri?), still awaiting the final means of officially reporting the bug, I can report that the simple line example is also directional, for example whether a line radiates outwards from the 0,0 origin in the positive direction, inward toward the origin and their complements in the negative direction along the x-axis (outward and inward).  Haven't finished varying along the y-axis or at angles, but the following code will allow others to investigate on their own. ''' LineDemo2.py Demonstrates errors in calculating various properties for polylines ''' import math import arcpy pnt = arcpy.Point() polylines = [] #vals = [0.001, 0.01, 0.1, 1.0, math.sqrt(2.0), 2.0] vals = [0.001, 0.01, 1.0 ] for a_val in vals:   polylines.append( [[0.0,0.0],[a_val,0.0]] )  #positive out   polylines.append( [[0.0,0.0],[-a_val,0.0]] ) #negative out   polylines.append( [[a_val,0.0],[0.0,0.0]] )  #positive in   polylines.append( [[-a_val,0.0],[0.0,0.0]] ) #negative in array = arcpy.Array() polys = [] for i in range(len(polylines)):   a_poly = polylines   print "\n", " Coordinates: ", a_poly   for pair in a_poly:     pnt.X = pair[0]     pnt.Y = pair[1]     array.add(pnt)     print " X: %20.16f  Y: %20.16f" % (pnt.X, pnt.Y)   poly = arcpy.Polyline(array)   cal_leng = poly.length   real_leng = max(abs(a_poly[1][0]), abs(a_poly[0][0]))   diff_leng = cal_leng - real_leng   print "Polyline properties:"   print " length:        %20.16f" % (cal_leng)   print " true length:   %20.16f" % (real_leng)   print " difference:    %20.16f" % (diff_leng)   print " percent error: %20.16f" % (diff_leng/real_leng * 100)   array.removeAll()   polys.append(poly)  #for further use results for 3 simple lengths, outward from the origin, inward towards the origin in both the positive and negative x directions >>> Coordinates:  [[0.0, 0.0], [1.0, 0.0]] X:   0.0000000000000000  Y:   0.0000000000000000 X:   1.0000000000000000  Y:   0.0000000000000000 Polyline properties: length:          1.0001220703125000 true length:     1.0000000000000000 difference:      0.0001220703125000 percent error:   0.0122070312500000 Coordinates:  [[0.0, 0.0], [-1.0, 0.0]] X:   0.0000000000000000  Y:   0.0000000000000000 X:  -1.0000000000000000  Y:   0.0000000000000000 Polyline properties: length:          1.0000000000000000 true length:     1.0000000000000000 difference:      0.0000000000000000 percent error:   0.0000000000000000 Coordinates:  [[1.0, 0.0], [0.0, 0.0]] X:   1.0000000000000000  Y:   0.0000000000000000 X:   0.0000000000000000  Y:   0.0000000000000000 Polyline properties: length:          1.0001220703125000 true length:     1.0000000000000000 difference:      0.0001220703125000 percent error:   0.0122070312500000 Coordinates:  [[-1.0, 0.0], [0.0, 0.0]] X:  -1.0000000000000000  Y:   0.0000000000000000 X:   0.0000000000000000  Y:   0.0000000000000000 Polyline properties: length:          1.0000000000000000 true length:     1.0000000000000000 difference:      0.0000000000000000 percent error:   0.0000000000000000 the gp group is on this and I will report when I can report ... View more

That means that a solution, if that was the case, hasn't been provided after all these years.  I am trying to demonstrate to students that the new powers of GIS and its incorporation of Python within its framework will make a simpler work environment for them...so I gave a simple demonstration of calculating the area of a unit square (etc) using the new "arcpy" environment and all its useful utilities, then I go on to show them how to do it in other environments, particular, pure Python...I had hoped to show that if you have ArcGIS installed use "arcpy" to get basic metric properties without the need to code your own procedures or use other modules (like Shapely etc)...unfortunately, those notes are burning on the bonfire outback. The more important issue is that there are many instances where feature geometry is manipulated (ie translated, rotated and scaled) to facilitate determinations of other geometric properties such as minimum area bounding circles, convex hulls, minimum area bounding ellipses, densifying features etc etc) since working in native projected coordinates is not appropriate (for a variety of reasons).  One would expect that any software would allow one to compute simple metrics for standard shapes centered at the origin and/or the minimum extent being at the origin.  If this is not the case (I never found an issue with this in AV3.x and/or calculations in 9.x that interfaced with Python), then a clear statement should be made that if you aren't using large coordinates, then you are going to get large errors.  ESRI has made the move to Python as your field calculator language de jour (VB Script as well) and ESRI strongly supports NumPy for a lot of the SA stuff (I presume, but I am not finished testing), so why can't I get a simple area for a unit square with the lower left extent at the origin from a module called "ARCPY", I thought, perhaps erroneously, that a synergy had developed...please someone, fix the issue, rather than defending it or bringing up the arcinfo past or call the module "ARCPYish"... rant over...thanks for the comments Chris PS Don't take offense, but this has serious ramifications for anyone that wants to use "a" GIS for any computational purposes.  The limitations have to be spelled out or addressed, there is no "silent" option By the way intersection of two polylines doesn't work for simply line features either ''' IntersectDemo.py Demonstrates errors in calculating various properties for polylines ''' import arcpy pnt = arcpy.Point() # #NOTE vary the epsilon parameter 1.0/8192 = 1.0/(2**13 #  Try 1.0/2.0**14 etc etc epsilon = 1.0/(2.0**13) polylines = [[[0.0,0.0], [1.0,0.0]], [[1.0 + epsilon, 0.0],[2.0,0.0]]] array = arcpy.Array() polys = [] for i in range(len(polylines)):   a_poly = polylines   print "\n", " Coordinates: ", a_poly   for pair in a_poly:     pnt.X = pair[0]     pnt.Y = pair[1]     array.add(pnt)     print " X: %20.16f  Y: %20.16f" % (pnt.X, pnt.Y)   poly = arcpy.Polyline(array)   print "Polyline properties:"   print " length %20.16f" % (poly.length)   print " centroid %s\n true centroid %s " % (poly.centroid, poly.trueCentroid)   print " first point %s\n last point %s " % (poly.firstPoint, poly.lastPoint)   print " extent %s " % (poly.extent)   #print " hull %s\n " % (poly.hullRectangle)   array.removeAll()   polys.append(poly) print "\n Does polyline 0 touch polyline 1?" print "answer ", polys[0].touches(polys[1]) inter_pnt_dist = polys[0].lastPoint.X - polys[1].firstPoint.X print "distance between last of 0 and 1st of 1 ", inter_pnt_dist results, vary the epsilon value to see different outcomes.  The results are wrong, the line segments should be separated by epsilon instead of 0.0 >>> Coordinates:  [[0.0, 0.0], [1.0, 0.0]] X:   0.0000000000000000  Y:   0.0000000000000000 X:   1.0000000000000000  Y:   0.0000000000000000 Polyline properties: length   1.0001220703125000 centroid 0.50006103515625 0 NaN NaN true centroid 0.50006103515625 0 NaN NaN first point 0 0 NaN NaN last point 1.0001220703125 0 NaN NaN extent 0 0 1.0001220703125 0 NaN NaN NaN NaN Coordinates:  [[1.0001220703125, 0.0], [2.0, 0.0]] X:   1.0001220703125000  Y:   0.0000000000000000 X:   2.0000000000000000  Y:   0.0000000000000000 Polyline properties: length   1.0000000000000000 centroid 1.5001220703125 0 NaN NaN true centroid 1.5001220703125 0 NaN NaN first point 1.0001220703125 0 NaN NaN last point 2.0001220703125 0 NaN NaN extent 1.0001220703125 0 2.0001220703125 0 NaN NaN NaN NaN Does polyline 0 touch polyline 1? answer  True distance between last of 0 and 1st of 1  0.0 ... View more