Downloadable benchmark tests and raw results from Jones et al., SeisIO: a fast, efficient geophysical data architecture for the Julia language, submitted to SRL.
- Install Julia ≥1.1 and Python ≥3.6. We used Julia v1.1.0 and Python v3.6.8 in the manuscript.
- (Optional) Get restricted files; see "Data and Resources" section in paper.
- Start Julia and run these commands:
using Pkg; Pkg.add("SeisIO"); using SeisIO; SeisIO_root = dirname(pathof(SeisIO))
. Save that location for the next step. - Place the restricted files in
SeisIO_root/../test/SampleFiles/Restricted/
.
- Start Julia and run these commands:
- SAC:
- Install SAC 101.6a from source
- Follow example template and instructions in
SAC/sac_shell.sh
- These runs produce screen output; dump or redirect to a temporary file
- Parse the temp file with grep -v to generate log files
SAC.read_sac.log
,SAC.read_suds.log
- If carefully checking memory, repeat the process with
time -v
and overwrite values inrun_sac_stats.jl
with the medians
- ObsPy:
- (Ubuntu 18.04) ensure pip is installed:
sudo apt install pip
- Modify the value of
SeisIO_path
inPython/ObsPy_bench1.py
if you need to run the "restricted" benchmarks. - Install ObsPy following these instructions
- Be sure to create and activate an ObsPy environment
- From the bash shell:
conda install pytest memory_profiler pyasdf
- Change directory to
Python/
. - Start python3.
- At the prompt:
exec(open("ObsPy_bench1.py").read())
- (Ubuntu 18.04) ensure pip is installed:
- SeisIO
- Start Julia in the root directory of this project.
- At the Julia prompt:
include("configure.jl") # configure Julia portion of project
include("benchmark_file_reads.jl") # run SeisIO benchmarks
include("benchmark_hist.jl") # generate plots for your own benchmarks