When using Quantile classification, each class is supposed to have the same number of pixels. However, with user's data, one class has a much larger number of pixels (~800,000) than other classes (~150,000).
上次发布: August 25, 2014ArcGIS for Desktop
漏洞 ID 编号
NIM081368
已提交
June 1, 2012
上次修改时间
June 5, 2024
适用范围
ArcGIS for Desktop
找到的版本
10.0
状态
Known Limit
经开发团队审核,已确定此问题与不受 Esri 控制的软件的已知限制有关。 问题的“其他信息”部分可能包含进一步说明。
附加信息
This input is highly skewed. The majority of the values are grouped into the first bin.
解决办法
1) Run Slice with equal area. Slice will calculate the appropriate areas based on the number of classes. Please note you may not end up with the exact number of classes specified due to the skewed data, however the class areas will be consistent as possible.2) Multiply by a large factor and convert to integer. Then run the reclassification on the resultant integer raster. This will provide an output that has very similar count values.from <a href="http://arcpy.sa" target="_blank">arcpy.sa</a> import *intRaster = Int(Raster("summer.berners.final.RSF.tif") * 1000000000)arcpy.BuildRasterAttributeTable_management(intRaster)run reclassify using raster input