Skip to content

HPC Repast with HDF5 modifications for SC15 Student Cluster Competition

License

Notifications You must be signed in to change notification settings

scamicha/SCC15HPCRepast

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

12 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

HPC Repast Installation instructions for SC15 Student Cluster Competition

The SC15 Student Cluster Competition will use a modified version of the zombie model for HPC Repast. This version (with modifications provided by Hoony Park of ORNL) allows for input from HDF5 files, as well as output to HDF5 on a cadence set by the user.

Installation

The installation of the modified zombie model follows the basic instructions presented in the INSTALL.txt files. However, there is an additional requirement of an HDF5 library. After installing a version of HDF5 > 1.8.14, modify the makefile to reference the installed HDF5 libraries. Following installation of all prerequisites (using the install.sh script if you wish) and all appropriate makefile modifications a

make hzombie_model

should build all of the modified zombie model code and its prerequisite HPC Repast libraries.

Configuration and Input

The configuration file model.props contains the variables for human and zombie input files. Here we describe the human files, but the zombie files are similar.

  • human.input.file the file containing the HDF5 input of the initial human distribution
  • human.output.file the file to hold the output of the human distribution at the given cadence
  • human.dataname the name of the HDF5 data block

In addition, the variable output.interval has been added to determine the frequency in timesteps at which the human and zombie population distributions are written to the output file.

Input can be generated by producing HDF5 files natively, or converting .csv files to HDF5 files. The python script csv2hdf.py in the src/scripts directory will convert a given .csv file to HDF5 format.

sample.csv is a 100x100 grid with a distribution of population members that you can experiment with.

About

HPC Repast with HDF5 modifications for SC15 Student Cluster Competition

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages