Use Con with two rasters

Question

Hello all

I am trying to write a python program that will create a new raster by condition of two rasters.

I can do it with the con function like this:

res = arcpy.sa.Con((Raster1 == 0 & Raster2 == 0),1,0)

For the con function you need to create a Raster() before you can use it and then save the results.

From the other hand arcpy.gp.con_sa  can get a path of tif as parameter and the final result name is parameter.

I could not find the way to put two rasters in the condition (it is usually "value == 0" of the raster in the new parameter.

Any ideas?

DanPatterson · Answer

Mody, if you want to bring the raster back in, check the optional parameters for rastertonumpyarray and numpyarraytoraster

RasterToNumPyArray—ArcGIS Pro | Documentation

RasterToNumPyArray (in_raster, {lower_left_corner}, {ncols}, {nrows}, {nodata_to_value})

NumPyArrayToRaster—ArcGIS Pro | Documentation

NumPyArrayToRaster (in_array, {lower_left_corner}, {x_cell_size}, {y_cell_size}, {value_to_nodata}, {mdinfo})

This allows you to maintain the spatial location and nodata values.

the lower left corner has examples in the help eg

lowerLeft = arcpy.Point(in_raster.extent.XMin, in_raster.extent.YMin)

mody_buchbinder · Answer

Hi Dan

Please read my previous question here: https://community.esri.com/t5/python-questions/what-is-the-way-to-call-spatial-analysis-tools-in-arcmap-from/td-p/1037005

The arcpy.sa.con_sa have different parameters and the first parameter is the input raster so you cannot gives two rasters.

I will look into the numpy

Thanks

DanPatterson · Answer

or a numpy lesson... verbose, showing that arcpy plays nice with numpy

a1 = arcpy.RasterToNumPyArray("c:/temp/a0.tif")
b1 = arcpy.RasterToNumPyArray("c:/temp/b0.tif")
whr = np.logical_and((a1==0), (b1==0))
out = np.where(whr, 1, 0)

a1 
array([[1, 0, 3, 0],
       [3, 2, 1, 1],
       [4, 3, 0, 3],
       [2, 0, 0, 1],
       [2, 4, 2, 1]])

b1 
array([[1, 1, 2, 4],
       [3, 1, 0, 3],
       [4, 2, 0, 1],
       [3, 4, 1, 2],
       [1, 0, 2, 0]])

whr 
array([[False, False, False, False],
       [False, False, False, False],
       [False, False,  True, False],
       [False, False, False, False],
       [False, False, False, False]])

out
array([[0, 0, 0, 0],
       [0, 0, 0, 0],
       [0, 0, 1, 0],
       [0, 0, 0, 0],
       [0, 0, 0, 0]])

my favorite raster goto 😉