#########
HPC4Stats
#########

Douglas Nychka, nychka@mines.edu and Dorit Hammerling,
dorith@ucar.edu

HPC4Stats: A framework for managing embarrassingly parallel data
analysis using R and the Rmpi package. 
Q21ZAQ~1 

Directories in this repository:

HPC4Stats.skeleton:

A tree of subdirectories (batch, data, output, src) that is the
recommended way to organize a computation/data analysis within
HPC4Stats. These directories are populated with README files that
explain the purpose of each directory and the files contained in
them. This skeleton is setup to do an example parallel computation and
see the batch subdirectory to run this example. You will need the
fields package installed to run this example, partly as a way to show
how packages are identified for computation and loaded.

HPC4StatsCMIP:

An example using a subset of climate model output and organized in the
same way as HPC4Stats.skeleton. The task is to fit a Generalized
Pareto distribution (GPD) to daily precipitation simulated by the NCAR
CCSM4 model. This is a historical run for the period 1955 01 01 to
1989 12 31. Here the GPD is fit by maximum likelihood separately for
every model grid box.  Note that to reduce the size of the data set
for the TEST example the full model output has been subsetted to be
just a grid of 21x51 locations and 7300 (~20 years) time points. This
is the R binary data object prExample.rda in the data
subdirectory.

Note: For the production example the user will need to install the
ncdf4 packageand to download or access the entire large netcdf file
(~2.6Gb):
"pr_day_CCSM4_historical_r1i1p1_19550101-19891231.nc". A link to this
file:

https://www.dropbox.com/s/80qbm1zjr7grx4i/pr_day_CCSM4_historical_r1i1p1_19550101-19891231.nc?dl=0

On NCAR's Cheyenne HPC system this file is available in the
directory:

/glade/p/CMIP/CMIP5/output1/NCAR/CCSM4/historical/day/atmos/day/r1i1p1/files/pr_20120202/

Acknowledgement: Development of this package was supported in part by
 the National Science Foundation Grant 1417857 and the National Center
 for Atmospheric Research.