How to Expand a raster with many zone values

[PatNussey]

I am using ArcMap 10.6.1 and I need to expand a large raster with thousands of zones by 1 cell using either the expand tool or a map algebra expression.  How can I write the expression to expand all zones in the input raster without entering each zone value separately.  Is there a way to write it so it expands zones from one number through to another? So for a range of raster values from the input raster. 

[DanPatterson_Retired]

I hope your zones aren't adjacent to one another since they might self obliterate.

You can copy and paste a zone list (not tried) using numpy.

import numpy as np
zones = np.arange(100)

zones
 
array([ 0,  1,  2,  3,  4,  5,  6,  7,  8,  9, 10, 11, 12, 13, 14, 15, 16,
       17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33,
       34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50,
       51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67,
       68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84,
       85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99])

# ---- just copy the list out like so
[ 0,  1,  2,  3,  4,  5,  6,  7,  8,  9, 10, 11, 12, 13, 14, 15, 16,
       17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33,
       34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50,
       51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67,
       68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84,
       85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99]‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍

[PatNussey]

It is a really big raster with over 700,000 zones, so this approach might be too cumbersome.  Is there a way to describe the zones by calling on the input raster's attribute table to iterate through the values? 

[DanPatterson_Retired]

This begs the question why are you expanding the zones?

Is it because it was a point file and you can't see them?

What do the zones represent? 

It is hard to picture an attribute represented by an integer as having 700,000 classes

[PatNussey]

It is a step in a model that was build with ArcInfo that does not seem to be accepted in ArcGIS Desktop.  The expand step follows a Region Group function on a raster to assign a unique ID to each contiguous raster area.  The expand step, I believe is to remove small gaps between region groups. The map algebra expression in the old model is: 

expand(wetrg, 1, table, %Workspace%\wetrg.vat)

Where wetrg is the region group output raster from the previous step.  The workspace has been specified in the environment settings. 

[DanPatterson_Retired]

So you have 700, 000 unique zones.  If you are trying to remove the fiddly-bits (aka, small areas), you normally query your zones on a size threshold as a precursor to using Nibble 

Nibble—Help | ArcGIS Desktop 

How Nibble works—Help | ArcGIS Desktop this too

[curtvprice]

It sounds like you are trying to replicate an existing workflow so I'm going to go back to Expand.

The 10x Expand syntax is:

Expand (in_raster, number_cells, zone_values)
‍‍‍‍‍

where zone_values is a python list of integers.

You can make a unique list from the VAT and use that variable with the Expand tool in Python like this. (If in_raster doesn't work, you may have to create a table view of the raster with MakeTableView and pass that in its place.)  

vlist = [v[0] for v in arcpy.sa.SearchCursor(in_raster, "VALUE")]
exp = arcpy.sa.Expand(in_raster, 1, vlist)‍‍‍‍
exp.save("exp_raster.tif")‍‍‍‍‍‍‍‍‍

Within Model Builder I think this is a good use case for the Calculate Value tool so you don't need to build a string representation of your many thousand integer values:

# Calculate Value expression
f(r"%Input raster%")

# Code block
def f(in_raster):
    vlist = [v[0] for v in arcpy.sa.SearchCursor(in_raster, "VALUE")]
    exp = arcpy.sa.Expand(in_raster, 1, vlist)‍‍‍‍
    exp.save("exp_raster.tif")‍‍‍‍‍‍‍‍‍
    return exp.catalogPath

# Data type
Raster Dataset‍‍‍‍‍‍‍‍‍‍‍‍

[PatNussey · Accepted solution]

Thank you for the suggestions.  I have found that the Euclidean Allocation tool seems to do the same thing if you set the distance to a multiple of the raster cell size. 

[curtvprice]

Good plan! You didn't mention you were expanding into NoData areas. If that's the case Euclidean Allocation is definitely what you needed. Either that or Nibble which uses a mask to select areas to be filled.