/pandas_bench

A simple benchmark for `pandas.DataFrame` being used in freqtrade.

Primary LanguagePythonGNU General Public License v3.0GPL-3.0

A simple benchmark for pandas.DataFrame being used in freqtrade.

We need to know whether PerformanceWarning is an indication of a significant slowdown or whether it can be safely ignored. Therefore we are collecting stats to make a decision.

Related issue on freqtrade

How to run

You will need docker and docker-compose

docker-compose up

Sharing results

Create a pull request with the updated CSV containing your values.

How to evaluate

  • Open eval.fods (works with LibreOffice)
  • Copy paste the contents of results.csv onto the results.csv sheet (don't replace the red text)
  • Switch to the second sheet (Called "Evaluation"), right click on the table and select "Refresh..."

You might need to extend the second table beside the first (pivot table) one if your CPU is hitherto unknown.

Current results

Generated with LibreOffice gen_img.sh

eval.jpg