Using Forest Based Classification and Regression Tool (in ArcGIS Pro 2.3.1) to predict to raster. Using PlaentScope imagery of 3 m spatial resolution to make predictions but getting the following error:
ERROR 110197: The number of raster cells to process 166861500 exceeds the maximum allowed 100000000.
I have used this tool previously for rasters covering a small area, and it was working perfectly fine. Now I am using large dataset and getting this error. Is there any way that we can increase this limit of Raster Cells to Process?
How Forest-based Classification and Regression works—ArcGIS Pro | ArcGIS Desktop
what options have you set and changed to see if it reduces the number of cells to process?
The only thing I have changed now is the raster size. Previously, I used raster covering an area of few hectares, but now I am trying to analyze rasters covering an entire county in Iowa. After this error, I analyzed one raster (covering relatively less area) at a time, and the tool worked fine. However, processing one raster at a time will add much extra time in processing.
Are you trying to train on the entire dataset? If so I would train on a sample or resampled version that isn’t so big.
For training, I am using a smaller dataset, and the training output is coming fine. The error appears once I run the "Predict to Raster" option.
Muhammad - I see you logged an Esri Support Services call on this question - that's great! Hopefully the analyst can discover the root cause of the error and/or create an Enhancement Request for Pro 2.x.