Fitting statistical distributions (wsim_fit)¶
The wsim_fit utility computes pixel-wise fits of statistical distributions.
Usage is as follows:
wsim_fit (--distribution=<dist>) (--input=<file>)... (--output=<file>)
--distributionspecifies the statistical distribution to be used. The following distributions are currently supported:gev: generalized extreme value distribution
--inputspecifies input files to be used. Multiple –input arguments can be provided, and each argument may refer to an individual file or a glob of files (e.g.,--input=precip_2017*.img)--outputspecifies the name of the output file.
All input values must be rasters of equal extent and resolution.
Program output is a single netCDF file, with global attributes indicating the variable fitted and distribution used, and float64 variables containing the fit parameters.
Example ncinfo output for a netCDF generated by wsim_fit follows:
root group (NETCDF3_CLASSIC data model, file format NETCDF3):
distribution: gev
dimensions(sizes): lon(720), lat(360)
variables(dimensions): float64 lon(lon), float64 lat(lat), float64 location(lat,lon), float64 scale(lat,lon), float64 shape(lat,lon)
groups: