Latest Contributions by DuncanHornby

I don't believe the variant of SQL used in ArcPro supports regular expressions in the Select by Attribute tool.  You can still use regular expressions but you need to step through your data with a search cursor, then update the selection set.  I provide the code below:   # Import modules import arcpy import re # Initialise objects layername = "p3" mapname = "ORN" fields = ["OBJECTID", "name"] oids = list() exp = '[A] \d+' # Get layer object p = arcpy.mp.ArcGISProject('current') m = p.listMaps(mapname)[0] l = m.listLayers(layername)[0] # Step through layer using regexp to test text in field with arcpy.da.SearchCursor(l, fields) as cursor: for row in cursor: oid = row[0] text = row[1] mo = re.search(exp, text) if mo is not None: oids.append(oid) # Set selection set for layer object l.setSelectionSet(oids, "NEW") ... View more

@gis_KIWI4 offers a solution that will guarantee that empty polygons are given a value, but which one? In their screen shot there are clearly two polygons adjacent to polygons with differing values. So as the code runs now it simply overwrites with the last visited polygon. If that is sufficient then OK but you might want to explore the Polygon Neighbours tool to help quantify the importance of adjacency by shared boundary length? ... View more

Reckon you could add a row to your table indicating level of supporting for source code encryption? ... View more

Just want to add my experience to the mix. When ever python breaks, notebook blows up etc etc I often find that the source of my problem is the fact I had installed Deep learning libraries for ArcGIS Pro. I have to use the Windows Add/Remove to remove the deep learning packages and often follow that up with a repair on ArcPro. That often fixes the issue. Something to explore... ... View more

OK now I see the problem you are asking about.  My experience when such tools fail to generate expected answers, in your case two first order streams must increment to a 2nd order when they come together, is that the under lying DEM was not corrected for hydrological analysis. Typically a  first step is to pass the DEM through the fill sinks tool before computing flow direction and or flow accumulation rasters.  My guess is that section of the DEM has not passed through the fill sinks tool?  I would run the DEM through that to see if it corrects it. ... View more

What you are showing is the raster attribute table of your stream order raster, you get that with integer rasters. Its simply the count of the number of pixels for the pixel value in the entire raster. So in your example you have 42340 pixels with a value of 1 and in that dataset 1 means a headwater stream. You don't have 42340 headwater streams you need to be counting the segment number to determine that. ... View more

Without your data or seeing how you ran the tool its just speculation, I would explore if your data has changed in some way, maybe the data you are joining is multi-part? ... View more

I want to compliment this post with further evidence of the 3.2 field mapping in the Export Table tool as being buggy.  If the input is a Table or FeatureClass then if you change the output field type properties (e.g. the field length of a text field say from 8000 to 255) it remembers it. If the input is a Feature Layer or Table View then any property changes of a field in the field map are instantly forgotten when you OK the field mapping dialog. It makes building models in model builder impossible when you need to use the field mapping, this needs to be fixed fast. ... View more

I think the nature of the tools are quite different, merge is simply bringing datasets together, dissolve or union is a geometric operation.  So to combine these disparate operations into a single tool is undoubtedly problematic.  So chaining the tools as you do now is the most robust solution. You can do this in model builder and expose it as a tool. You then get the experience of a single tool interface but behind the scenes its running the tools in sequence. Suggest you explore model builder. ... View more

Looking forward to part two! ... View more

This should get you going, this is how I would have done it.  Code assumes single part and a line with a minimum of 4 vertices: import arcpy lyrLine = "lines" lstRes = list() with arcpy.da.SearchCursor(lyrLine, "SHAPE@") as cursor: for row in cursor: geom = row[0] # A polyline if geom.pointCount >= 4: arr =geom.getPart(0) p1 = arr[0] # first p2 = arr[1] # second p3 = arr[-2] # second from last p4 = arr[-1] # last s1 = str(p1.X) + "," + str(p1.Y) s2 = str(p2.X) + "," + str(p2.Y) s3 = str(p3.X) + "," + str(p3.Y) s4 = str(p4.X) + "," + str(p4.Y) lstRes.append([s1,s2,s3,s4]) print(lstRes)   ... View more

Have you tried the indexing which you suggest as that is what I would have done? ... View more

I can't replicate your problem, how are you running this script? ... View more

Yes that's the correct tool, but your where clause is incorrect ToDate is not being referenced correctly. Read the help for this tool and explore the samples to understand how to build the clause correctly.   ... View more

You are using the wrong Select tool, you want to be using the Select by Attributes tool. ... View more