Terrible Directions

• Run "query.py" on accounting.chtc.wisc.edu to query for data in a range

    $ python3 query.py > YYYY-MM-DD_Ndays.csv
  

• R script will filter down to jobs that ran on prioritized nodes (we think)

    $ Rscript scratch.R
  

• Python creates nice summary tables based on user + node

    $ python3 scrach.py
  

Changes Christina would make

• Add more arguments (right now you have to manually edit the filenames / dates)
• Turn the R script into a Python script
• Maybe include backfill for prioritized nodes, not just prioritized? to calculate prioritized vs backfill