cdscan speed about cdms HOT 6 OPEN

@pochedls did you try the -i -j and --time-linear options?

from cdms.

@doutriaux1 - I was worried about using some of these options, because I didn't know if linearity was always a good option. It seemed likely that some models may have done weird things that would breaj that assumption.

The -j option doesn't speed things up. The -i and --time-linear options give me an error (not sure if it is my syntax or a bug).

time cdscan -x test.xml --time-linear /p/css03/cmip5_css02/data/cmip5/output1/MIROC/MIROC5/historical/mon/ocean/Omon/r2i1p1/vo/1/*.nc
Traceback (most recent call last):
File "/export_backup/pochedley1/bin/anaconda2/envs/cdat81/bin/cdscan", line 1840, in
main(sys.argv)
File "/export_backup/pochedley1/bin/anaconda2/envs/cdat81/bin/cdscan", line 837, in main
ttzero = float(targlist[0])
ValueError: could not convert string to float: '/p/css03/cmip5_css02/data/cmip5/output1/MIROC/MIROC5/historical/mon/ocean/Omon/r2i1p1/vo/1/vo_Omon_MIROC5_historical_r2i1p1_185001-185012.nc'

from cdms.

@dnadeau4 @doutriaux1 this is a key enhancement request, can we get it put on the "PRIORITY" queue?

from cdms.

I was working on this, but thought that it was no longer necessary to speed it up.

from cdms.

@dnadeau4 cdscan is a full workhorse, and any speedups/optimizations/bugfixes that can be implemented would be key

from cdms.

Adding one vote for putting cdscan enhancement on the "PRIORITY" queue..

from cdms.