I'm trying to set the symbology on several hundred .tiff raster files.
I need to customize the statistics and would like to symbolize using a stretch MinMax function on each raster.
So I've set up the custom statistics in the raster function as below:
Are they singleband tiff files?
Are the data integer or floating point?
What are the expected min and max's?
Do all rasters cover basically the same min and max range?
Singe-band 32-bit floats
mins and maxes somewhere in the ranges from 0-1000 of different meteorological indicators
rasters in sets of four cover the same min/max ranges, 420 maps total
Yep but I've set my Raster and Imagery defaults to a different colorramp, should have mentioned. It would typically default to grayscale, but still be classified instead of stretch
the default resampling method is bilinear and the default stretch is minimum-maximum.
That is why I wonder why you mention ColorMap instead of ColorRamp
I refer to the ColorMap raster function, which has colorRamp as an option within it
I have found a workaround to this by attaching a processing template which calcs the statistics (and does nothing else) and then set the rendering defaults to my desired symbology within ArcGIS Pro.
But I'm looking for a solution to this symbology issue with raster functions because I'd like to publish it as an image service and have the correct symbology already applied instead of using client-side rendering to change the vis (which is was I'm doing now in Esri JS API)
Did you ever find a solution Giannalogy?
I'm trying to publish an image service, with a predefined stretch/colour ramp and came up with the same workflow.
I've built the Raster Function template, but cannot get the Colormap and Stretch tools to work together - so I end up with the classified raster, same as you.
Thanks. I can see that we are trying to apply two different visual functions, so they can't be chained. But the stretch function should allow you to apply a colour ramp.
It's also supposed to be possible from the Symbology tool:
But this method doesn't work for me either. It manages to apply the correct color ramp, but seems to now think it's an RGB raster, and I lose access to the original band data.
Time to lodge a case for this method - or try and hack the Raster Function xml :\