topic Raster divide error in ArcGIS Spatial Analyst Questions https://community.esri.com/t5/arcgis-spatial-analyst-questions/raster-divide-error/m-p/438772#M6177 <HTML><HEAD></HEAD><BODY><SPAN>I am trying to calculate a ratio of 48 binary rasters. I have written an AML to perform the analysis but there are a a few grids that the resulting float raster exceeds 32-bit addressing space. It is my understanding that ArcMap/Spatial Analyst can address larger rasters so I calculated the integer rasters (using focalsum) and tried to calculate the ratio in both raster calculator and Single Output Map Algebra (ArcToolbox). Both methods yield a float raster of a single value (0.5). I can subset the data to a smaller raster and get the correct results and I can recreate the problem on the other large datasets. Any ideas on why this is happening and a workaround that does not require subsetting the data?</SPAN><BR /><BR /><SPAN>Here is my statement for calculating the ratio (works on smaller rasters):</SPAN><BR /><SPAN>float((float([sum1]) / float([sum2]))) </SPAN><BR /><BR /><SPAN>To calculate the rasters used above, starting with a binary raster:</SPAN><BR /><SPAN>tmp1 = CON(ingrd == 0, 1, 0)</SPAN><BR /><SPAN> sum1 = FOCALSUM(ingrd, circle, 1000)</SPAN><BR /><SPAN>&nbsp; tmp3 = FOCALSUM(tmp1, circle, 1000)</SPAN><BR /><SPAN>sum2 = sum1 + tmp3</SPAN></BODY></HTML> Tue, 11 May 2010 21:14:07 GMT JeffreyEvans 2010-05-11T21:14:07Z Raster divide error https://community.esri.com/t5/arcgis-spatial-analyst-questions/raster-divide-error/m-p/438772#M6177 <HTML><HEAD></HEAD><BODY><SPAN>I am trying to calculate a ratio of 48 binary rasters. I have written an AML to perform the analysis but there are a a few grids that the resulting float raster exceeds 32-bit addressing space. It is my understanding that ArcMap/Spatial Analyst can address larger rasters so I calculated the integer rasters (using focalsum) and tried to calculate the ratio in both raster calculator and Single Output Map Algebra (ArcToolbox). Both methods yield a float raster of a single value (0.5). I can subset the data to a smaller raster and get the correct results and I can recreate the problem on the other large datasets. Any ideas on why this is happening and a workaround that does not require subsetting the data?</SPAN><BR /><BR /><SPAN>Here is my statement for calculating the ratio (works on smaller rasters):</SPAN><BR /><SPAN>float((float([sum1]) / float([sum2]))) </SPAN><BR /><BR /><SPAN>To calculate the rasters used above, starting with a binary raster:</SPAN><BR /><SPAN>tmp1 = CON(ingrd == 0, 1, 0)</SPAN><BR /><SPAN> sum1 = FOCALSUM(ingrd, circle, 1000)</SPAN><BR /><SPAN>&nbsp; tmp3 = FOCALSUM(tmp1, circle, 1000)</SPAN><BR /><SPAN>sum2 = sum1 + tmp3</SPAN></BODY></HTML> Tue, 11 May 2010 21:14:07 GMT https://community.esri.com/t5/arcgis-spatial-analyst-questions/raster-divide-error/m-p/438772#M6177 JeffreyEvans 2010-05-11T21:14:07Z