lmcinnes/umap

UMAP Roadmap

Open

#15 建立於 2017年11月13日

在 GitHub 查看
 (36 留言) (22 反應) (0 負責人)Python (6,478 star) (751 fork)batch import
help wantednew feature

描述

A rough roadmap of things to be done for UMAP. Some of these tasks are easy, some are hard, and some require deeper knowledge of UMAP. Short and medium term tasks should be approachable for many people. Reply to this issue if you are interested in taking up any of them.

Short term items

  • Support for sparse matrix input
  • Add random seed as an user option
  • Support for cosine distance RP-trees
  • Allow non-RP-tree initialisation of NN-descent
  • Better document (via docstrings) all the support functions
  • "Custom" initialisation with a predefined positioning.

Medium term items

  • Generate notebook for basic usage demonstration
  • Generate notebook explaining parameter options and their effects
  • Set up CI and build a basic test suite
  • Start building basic documentation and integrate with readthedocs

Longer term items

  • Generate notebook for "How UMAP works"
  • Add code (and devise API(?)) for UMAP on general pandas dataframes
  • Add support for semi-supervised dimension reduction via UMAP
  • UMAP as a generative model (code + demo)
  • UMAP for text data (similar to word2vec)
  • A transform function for new previously unseen data (see issue #40)
  • Model persistence for UMAP models

No priority

  • GPU support for UMAP
  • Conda-forge UMAP package
  • Improve numba usage (better numba expertise required)
  • Concurrency via Dask for multicore and distributed support

貢獻者指南

技術棧
pythonnumpy
領域
machine learning
議題類型
feature
難度面向新貢獻者的預計實作難度，1 表示很小改動，5 表示專家級工作。
3
預計時間有經驗貢獻者完成調查、實作、測試並準備 pull request 的粗略時間範圍。
over 1 week
活動狀態議題目前的可參與程度：新鮮、活躍、陳舊、阻塞或等待維護者輸入。
active
清晰度議題是否清楚說明預期改動、驗收標準和下一步。
needs investigation
前置要求
understanding of UMAPPythonnumpynumba
新手友善度1-100 的估計分數，表示該議題對首次貢獻者的友善程度。
30
研究方向
This issue is a roadmap outlining multiple tasks for UMAP. The remaining open tasks include building documentation with readthedocs, creating a 'How UMAP works' notebook, adding support for pandas dataframes, semi supervised reduction, generative models, text data, transform for new data, and model persistence. To get started, focus on the medium term uncompleted task: 'Start building basic documentation and integrate with readthedocs'. This involves setting up Sphinx documentation and deploying to readthedocs. Look at existing notebooks in the repository for style and content. Alternatively, the short term item 'Generate notebook explaining parameter options and their effects' is also unfinished (not checked). Review the project structure and existing comments for guidance.
UMAP Roadmap · lmcinnes/umap#15 | Good First Issue