Very drafty!!! TODO: add more links from notes and structure it better

sklearn

• tiny-sklearn Minimal implementation of sklearn. Good to get general idea. https://github.com/qinhanmin2014/tiny-sklearn

• timeseriescv split by time for sklearn https://github.com/sam31415/timeseriescv

• trees for sklearn https://github.com/scikit-garden/scikit-garden

• sklearn extensions https://github.com/koaning/scikit-lego

• sklearn searchgrid syntax sugar https://github.com/jnothman/searchgrid

• hyperopt for sklearn https://github.com/hyperopt/hyperopt-sklearn

• sklearn pipeline extensions https://github.com/jem1031/pandas-pipelines-custom-transformers

https://github.com/Kgoetsch/sklearn_pipeline_enhancements

• sklearn extensions* https://github.com/scikit-learn-contrib/scikit-learn-extra

• sklern SLEP https://scikit-learn-enhancement-proposals.readthedocs.io/en/latest/

• sklearn extensions https://github.com/rasbt/mlxtend

frameworks

• Kedro by quantumblack, cookiecutter-data-science https://github.com/quantumblacklabs/kedro

https://github.com/drivendata/cookiecutter-data-science

• MLFlow https://github.com/mlflow/mlflow

• DVC - data versioning https://github.com/iterative/dvc

• Hangar - data versioning https://github.com/tensorwerk/hangar-py

SQL

• SQL on GPU https://blazingsql.com/

AutoML - single table

• NEURAXLE, TO_CHECK Let your pipeline steps have hyperparameter spaces https://github.com/Neuraxio/Neuraxle

• LALE Sklearn-like pipelines on steroids regarding hyper-parameters and type checking. Interoperability. IBM https://github.com/IBM/lale

• FKLEARN Functional sklearn by NuBank. Feature names. Spatio-temporal split during validation. https://github.com/nubank/fklearn

• DABL Simple but powerful ML. Really nice EDA. By amueller from sklearn. https://github.com/amueller/dabl

• hyperparameter_hunter https://github.com/HunterMcGushion/hyperparameter_hunter

• visualization for automl https://github.com/HDI-Project/ATMSeer

• very high-level, by uber https://github.com/uber/ludwig

AutoML - multiple tables

• get.ML Closed-source, trial, C++, incremental updates https://www.get.ml/

Pandas, data wrangling

• Sets handling for pandas https://github.com/Florents-Tselai/pandas-sets

• Pandas summary https://github.com/mouradmourafiq/pandas-summary

• Pandas column type detection https://github.com/tahaceritli/ptype-dmkd

• Pandas EDA https://github.com/pandas-profiling/pandas-profiling

• CleverCSV Dealing with messy CSV https://github.com/alan-turing-institute/CleverCSV

• dirty categorical variables https://github.com/dirty-cat/dirty_cat/

• great-expectations Testing pipelines and pandas https://github.com/great-expectations/great_expectations

• defensive pandas https://github.com/engarde-dev/engarde

• Intake Data directories. Compare with Kedro-intake https://github.com/intake/intake

• data cleaning https://github.com/ericmjl/pyjanitor

• MetaFlow NetFlix, AWS https://metaflow.org/

• serverless pipelines https://nuclio.io/

• feature store https://github.com/gojek/feast

Time series

• Matrix-profile Really interesting approach https://github.com/target/matrixprofile-ts

• sklearn inspired https://github.com/EthanRosenthal/skits

• fast.ai timeseries https://github.com/tcapelle/TimeSeries_fastai

Tools - ML (visualization)

• By uber - ml debug https://github.com/uber/manifold

• By google - ml debug https://github.com/pair-code/what-if-tool

• Compare models https://github.com/JohannesBuchner/UltraNest

• Built AI apps fast https://www.streamlit.io/

Visualization

https://github.com/plotly/plotly_express

https://github.com/gyli/PyWaffle

https://github.com/holoviz/hvplot - pandas

Arrays, big data

https://github.com/zarr-developers/zarr-python

https://github.com/dask/dask

https://github.com/scikit-hep/awkward-array

https://github.com/ibis-project/ibis

Tools - general

• Add structure to your commits https://github.com/commitizen/cz-cli

• git and jupyter https://github.com/jupyter/nbdime

• parametrizaton of jupyter https://github.com/nteract/papermill

• jupyter into dashboard https://github.com/voila-dashboards/voila

• Git PL https://rogerdudler.github.io/git-guide/index.pl.html

• Black python code formatter https://github.com/psf/black

• tabulate ... https://github.com/gregbanks/python-tabulate

ML - Reinforcement Learning

• Really good intro https://mpatacchiola.github.io/blog/
• ReAgent

RL by FB, PyTorch based https://github.com/facebookresearch/ReAgent

• MuZero https://github.com/werner-duvaud/muzero-general

NLP, NLG etc.

• TEXAR https://github.com/asyml/texar

• PPLM Text generation by uber https://eng.uber.com/pplm/

• fast-bert Inspired by fast.ai https://github.com/kaushaltrivedi/fast-bert

Casual inference

• CausalLML By uber. Lifting modeling. https://github.com/uber/causalml

• Churn - MARKETING https://github.com/CamDavidsonPilon/lifetimes

Other

• Notes for professional book series https://books.goalkicker.com/

• Deep-quant. Fundamental data https://github.com/euclidjda/deep-quant

• Google data search https://datasetsearch.research.google.com/