SDR model ValueError

Dear all,
I'm working with SDR model, InVEST 3.3.2 on Linux Ubuntu with my own dataset.
I faced the following error message when I tried to run it:

ValueError: There was not a value for at least the following codes [-10000] for this file /home/elena/LAVORI/AssegnoRicerca/Corsi/InVEST/ProvaInVEST/SDR_BO/aligned_lulc.tif.
Nodata value is: -9999

I attach the log file.

I think it's a problem of LULC alignment, but my input data are all in the same Reference System (ETRS89 UTM 32N) with the same spatial resolution, with null vaules as -9999. -10000 looks like a rounding of my null values.
If you need the original data I can attach them.

Thank you for your helpfulness.

Elena




Tagged:

Comments

  • RichRich Administrator, NatCap Staff
    Hi Elena, it looks like what you said... You have a -10000 value in your LULC map, but it probably was supposed to be -9999 to represent nodata values.  Any chance you can reclassify those -10000s to -9999 or change your nodata value of your raster to -9999?
  • Hi Rich,
    No data value are already -9999. I can't reclassify these -10000 because they are generated by the SDR procedure doing the alignment.
    I'm wondering about putting a -10000 code in the biophysical table with C and P value -9999, in your opinion would it work? SDR would consider them as null value?
  • RichRich Administrator, NatCap Staff
    Ahh that's strange -- SDR changing nodata values would be a bug we should patch.  Would you mind dropboxing your data to me at richsharp@stanford.edu so I can try to recreate the issue and patch the issue for a future release?

    And sure, you can add an entry in your biophysical table for -10000, but I'm worried the extent of it.  If it intersects with parts of the DEM that are supposed to be masked out then you might get some weird results.
  • By adding a new entry in the biophysical table for -10000 it works. I can't add C and P value of -9999, because C and P have to be floating and between 0 and 1. Thus I put equal 0.
    I'm dropboxing you my original data.
Sign In or Register to comment.