microsoft / microsoft/qlib

[Feedback Request] Bodo DataFrames for data transformations

Open
#2,017 0 comments 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

enhancement
Dominant language
Python
Stars
48.7k
Forks
7.7k
PR merge metrics
No merged PRs in 30d

Description

🌟 Feature Description

Integrate Bodo Dataframe's or compiler into qlib as an optional execution backend. Bodo Dataframe's are a drop in replacement for Pandas dataframes and offer expression tree optimization with lazy evaluation and parallel execution that can scale to multiple nodes. It automatically fallsback to Pandas if an unsupported operation is encountered. Bodo's compiler is built on Numba and automatically parallelizes numpy/pandas/python code across one or more nodes. This could accelerate qlib expression execution or data transformations

Motivation

  1. Application scenario
    Accelerating/Scaling qlib data transformations, potentially to multiple nodes
  2. Related works (Papers, Github repos etc.):
    https://github.com/bodo-ai/bodo

Alternatives

There are other options for high performance/parallel dataframes such as polars and pyspark but polars isn't pandas compatible increasing the integration effort and pyspark can be complicated to configure and deploy and doesn't offer competitive performance in our experience.

Additional Notes

My main purpose in creating this issue is gauging if there's interest and if so the best places to start looking for a POC that I would work on. Thanks!

Contributor guide

No contributing guide indexed for this repository

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Research direction

No qlib files, tests, or entry points are named. Start by reviewing the Bodo repository and qlib's data-transformation and expression-execution areas, then define a small proof of concept, its optional-backend boundary, and how Pandas fallback would be validated.

Written by the indexing model from the issue text.

Assessment

Tech stack
numpy, pandas, python
Domain
data-engineering, performance
Issue type
Feature
Difficulty
5/5
Estimated time
Over a week
Activity status
Stale
Clarity
Needs clarification
Newbie friendliness
25/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.