>>
>
>NCL will not recognize "NaN", even as a missing value. You need to have
>an actual number like "1e36" or "-9999" for the missing value.
>
>If you get "NaN"s in your netCDF data file, then you might want to go
>back to the source that created the file, and make sure that you write
>out real values instead of NaNs.
I have never seen "Nan" used as a missing value. However, it
can be changed 'on-the-fly'
sst = f->ssta
sst@_FillValue = -999.9 ; whatever
sst = sst = mask(sst, sst.eq."NaN", False)
print (sst)
All NaN will be changed to -999.9.
Why does the sst.eq."NaN" work?
It is because sst is of type float amd float is coercible to
a variable of type string.
Dennis Shea
http://ngwww.ucar.edu/ngdoc/ng4.1/ref/ncl/NclDataTypes.html#Coercion
_______________________________________________
ncl-talk mailing list
ncl-talk@ucar.edu
http://mailman.ucar.edu/mailman/listinfo/ncl-talk
This archive was generated by hypermail 2b29 : Mon Jul 29 2002 - 16:40:15 MDT