Directories in this archive correspond to a dataset for a particular \
heterogeneous computing system and workload.

E1 -- 9 machine types in the system
E2 -- 5 machine types in this system

The data is either in CSV or JSON format.  The file extensions csv and json \
are used to distringuish the formats.

Input Files:
eta.csv -- operations per task
eta variance.csv -- variance of the eta matrix (element-wise) 
psi.csv -- energy per operation [J]
psi variance.csv -- variance of psi matrix (element-wise)
tau.csv -- time per operation [s]
tau variance.csv -- variance of tau matrix (element-wise)
tasks.csv -- describes the task types, the header describes the details
machines.csv -- describes the machine types, the header describes the \
details
params.json -- Master parameter file that contains the various scalar \
parameters not in the previous files

Output Files Directories:
scaling -- data for the monte carlo runs for the inital scalbility and run \
time data
compare -- the data used for the comparison with the heuristics (it's just \
one run).
paretoRP -- the pareto from data for the reward rate and power bi-objective \
problem
feasibleRP -- feasible region data for the reward rate and power \
bi-objective problem
