arcgis cellstatistics (range) error

500
5
Jump to solution
07-18-2017 07:58 PM
changkx
New Contributor

I want to calculate the range sequence, the Python script as follow:

  1. #ndvi_R = CellStatistics(RList, "RANGE", "DATA")the results are 0, but when I run as follows, it's correct. 
  2. ndvi_MAX = CellStatistics(RList, "MAXIMUM", "DATA")
    ndvi_MIN = CellStatistics(RList, "MINIMUM", "DATA")
    ndvi_R = ndvi_MAX - ndvi_MIN
0 Kudos
1 Solution

Accepted Solutions
DanPatterson_Retired
MVP Esteemed Contributor

this weirdness with cell statistics has been reported before

https://community.esri.com/thread/179829

I will look at the rasters, but I suspect it is a nodata issue, an extent issue or similar related environment properties

EDIT

  1. I can confirm that the problem remains if the inputs are 32bit tiffs or 32 bit esri grids
  2. Your workaround does the the job.
  3. I haven't gone to the effort of converting to a lower bit level... that could be a task for someone else.
  4. your band 3 is pretty strange... calculate statistics, try viewing without a stretched symbology and have a look at its histogram
  5. try scaling the values to integer (ie   Int('band1'*100) ) then determining the range

I would file a case with tech support, especially in light of esri's default grid format being tiffs in recent versions.

View solution in original post

5 Replies
DanPatterson_Retired
MVP Esteemed Contributor

check your rlist since 0 is expected under one circumstance.  Your workaround is fine.. perhaps you have a nodata value in one of the rasters... too hard to tell from your description

For statistic types Maximum, Minimum, Mean, Median, Majority, Minority and Sum, if a single raster is used as the input, the output cell values will be the same as the input cell values. For Range and STD the output cell values will all be 0, and for Variety, all 1.

changkx
New Contributor
  • Thanks for your help, I calculate one RList as follows alone, but the results are still 0. If it exists a nodata value, the parameter ‘DATA’ was explained in the arcgis help. 

    • DATA —At the processing cell location, if any of the input rasters has NoData, then that NoData value will be ignored. The statistics will be computed by only considering the cells with valid data. This is the default.

     

  • RList
  • I have attached the RList:ndvi_R.zip - Google Drive 
0 Kudos
DanPatterson_Retired
MVP Esteemed Contributor

this weirdness with cell statistics has been reported before

https://community.esri.com/thread/179829

I will look at the rasters, but I suspect it is a nodata issue, an extent issue or similar related environment properties

EDIT

  1. I can confirm that the problem remains if the inputs are 32bit tiffs or 32 bit esri grids
  2. Your workaround does the the job.
  3. I haven't gone to the effort of converting to a lower bit level... that could be a task for someone else.
  4. your band 3 is pretty strange... calculate statistics, try viewing without a stretched symbology and have a look at its histogram
  5. try scaling the values to integer (ie   Int('band1'*100) ) then determining the range

I would file a case with tech support, especially in light of esri's default grid format being tiffs in recent versions.

changkx
New Contributor

Thanks. I have checked my RList seriously and removed the 'X_3_3.tif', but the problem still exists. Finally, I scaled the values to integer (Int("X_2_3.tif" * 100)) then determining the range, it works. Thank you for your patience and help.

0 Kudos
DanPatterson_Retired
MVP Esteemed Contributor

glad it worked... I suspect this issue will arise again... bookmarking

0 Kudos