Awesome Data Science with Python

A curated list of awesome resources for practicing data science using Python, including not only libraries, but also links to tutorials, code snippets, blog posts and talks.

https://github.com/r0f1/datascience

Core

pandas - Data structures built on top of numpy.
scikit-learn - Core ML library.
matplotlib - Plotting library.
seaborn - Data visualization library based on matplotlib.
pandas_summary - Basic statistics using DataFrameSummary(df).summary().
pandas_profiling - Descriptive statistics using ProfileReport.
sklearn_pandas - Helpful DataFrameMapper class.
missingno - Missing data visualization.
rainbow-csv - Plugin to display .csv files with nice colors.

Environment and Jupyter

General Jupyter Tricks
Fixing environment: link
Python debugger (pdb) - blog post, video, cheatsheet
cookiecutter-data-science - Project template for data science projects.
nteract - Open Jupyter Notebooks with doubleclick.
papermill - Parameterize and execute Jupyter notebooks, tutorial.
nbdime - Diff two notebook files, Alternative GitHub App: ReviewNB.
RISE - Turn Jupyter notebooks into presentations.
qgrid - Pandas DataFrame sorting.
pivottablejs - Drag n drop Pivot Tables and Charts for jupyter notebooks.
itables - Interactive tables in Jupyter.
jupyter-datatables - Interactive tables in Jupyter.
debugger - Visual debugger for Jupyter.
nbcommands - View and search notebooks from terminal.
handcalcs - More convenient way of writing mathematical equations in Jupyter.

Pandas Tricks, Alternatives and Additions

Pandas Tricks
Using df.pipe() (video)
pandasvault - Large collection of pandas tricks.

modin - Parallelization library for faster pandas DataFrame.
vaex - Out-of-Core DataFrames.
pandarallel - Parallelize pandas operations.
xarray - Extends pandas to n-dimensional arrays.
swifter - Apply any function to a pandas dataframe faster.
pandas_flavor - Write custom accessors like .str and .dt.
pandas-log - Find business logic issues and performance issues in pandas.
pandapy - Additional features for pandas.

Helpful

tqdm - Progress bars for for-loops. Also supports pandas apply().
icecream - Simple debugging output.
loguru - Python logging.
pyprojroot - Helpful here() command from R.
intake - Loading datasets made easier, talk.

Extraction

textract - Extract text from any document.
camelot - Extract text from PDF.

Big Data

spark - DataFrame for big data, cheatsheet, tutorial.
sparkit-learn, spark-deep-learning - ML frameworks for spark.
koalas - Pandas API on Apache Spark.
dask, dask-ml - Pandas DataFrame for big data and machine learning library, resources, talk1, talk2, notebooks, videos.
dask-gateway - Managing dask clusters.
turicreate - Helpful SFrame class for out-of-memory dataframes.
h2o - Helpful H2OFrame class for out-of-memory dataframes.
datatable - Data Table for big data support.
cuDF - GPU DataFrame Library.
ray - Flexible, high-performance distributed execution framework.
mars - Tensor-based unified framework for large-scale data computation.
bottleneck - Fast NumPy array functions written in C.
bolz - A columnar data container that can be compressed.
cupy - NumPy-like API accelerated with CUDA.
petastorm - Data access library for parquet files by Uber.
zappy - Distributed numpy arrays.

Command line tools, CSV

ni - Command line tool for big data.
xsv - Command line tool for indexing, slicing, analyzing, splitting and joining CSV files.
csvkit - Another command line tool for CSV files.
csvsort - Sort large csv files.
tsv-utils - Tools for working with CSV files by ebay.
cheat - Make cheatsheets for command line commands.

Classical Statistics

Statistical Tests and Packages

pingouin - Statistical tests.
scipy.stats - Statistical tests.
researchpy - Helpful summary_cont() function for summary statistics (Table 1).
scikit-posthocs - Statistical post-hoc tests for pairwise multiple comparisons.
Bland-Altman Plot - Plot for agreement between two methods of measurement.
ANOVA, Tutorials: One-way, Two-way, Type 1,2,3 explained.

Visualizations

Null Hypothesis Significance Testing (NHST) and Sample Size Calculation
Correlation
Cohen's d
Confidence Interval
Equivalence, non-inferiority and superiority testing
Bayesian two-sample t test
Distribution of p-values when comparing two groups
Understanding the t-distribution and its normal approximation

Talks

Inverse Propensity Weighting
Dealing with Selection Bias By Propensity Based Feature Selection

Texts

Greenland - Statistical tests, P values, confidence intervals, and power: a guide to misinterpretations
Lindeløv - Common statistical tests are linear models
Chatruc - The Central Limit Theorem and its misuse
Al-Saleh - Properties of the Standard Deviation that are Rarely Mentioned in Classrooms
Wainer - The Most Dangerous Equation
Gigerenzer - The Bias Bias in Behavioral Economics
Cook - Estimating the chances of something that hasn’t happened yet

Frameworks

scikit-learn - General machine learning framework.
h2o - Machine learning framework.
caffe - Deep learning framework, pretrained models.
mxnet - Deep learning framework, book.

Exploration and Cleaning

Checklist.
littleballoffur - Sampling from graphs.
janitor - Clean messy column names.
impyute - Imputations.
fancyimpute - Matrix completion and imputation algorithms.
imbalanced-learn - Resampling for imbalanced datasets.
tspreprocess - Time series preprocessing: Denoising, Compression, Resampling.
Kaggler - Utility functions (OneHotEncoder(min_obs=100))
pyupset - Visualizing intersecting sets.
pyemd - Earth Mover's Distance, similarity between histograms.

Train / Test Split

iterative-stratification - Stratification of multilabel data.

Feature Engineering

Talk
sklearn - Pipeline, examples.
pdpipe - Pipelines for DataFrames.
scikit-lego - Custom transformers for pipelines.
skoot - Pipeline helper functions.
categorical-encoding - Categorical encoding of variables, vtreat (R package).
dirty_cat - Encoding dirty categorical variables.
patsy - R-like syntax for statistical models.
mlxtend - LDA.
featuretools - Automated feature engineering, example.
tsfresh - Time series feature engineering.
pypeln - Concurrent data pipelines.
feature_engine - Encoders, transformers, etc.

Feature Selection

Talk
Blog post series - 1, 2, 3, 4
Tutorials - 1, 2
sklearn - Feature selection.
eli5 - Feature selection using permutation importance.
scikit-feature - Feature selection algorithms.
stability-selection - Stability selection.
scikit-rebate - Relief-based feature selection algorithms.
scikit-genetic - Genetic feature selection.
boruta_py - Feature selection, explaination, example.
linselect - Feature selection package.
mlxtend - Exhaustive feature selection.
BoostARoota - Xgboost feature selection algorithm.
INVASE - Instance-wise Variable Selection using Neural Networks.

Dimensionality Reduction

Talk
prince - Dimensionality reduction, factor analysis (PCA, MCA, CA, FAMD).
sklearn - Multidimensional scaling (MDS).
sklearn - t-distributed Stochastic Neighbor Embedding (t-SNE), intro. Faster implementations: lvdmaaten, MulticoreTSNE.
FIt-SNE - Fast Fourier Transform-accelerated Interpolation-based t-SNE.
umap - Uniform Manifold Approximation and Projection, talk, explorer, explanation.
sleepwalk - Explore embeddings, interactive visualization (R package).
scikit-tda - Topological Data Analysis, paper, talk, talk.
mdr - Dimensionality reduction, multifactor dimensionality reduction (MDR).
sklearn - Truncated SVD (aka LSA).
ivis - Dimensionality reduction using Siamese Networks.
trimap - Dimensionality reduction using triplets.

Visualization

All charts, Austrian monuments.
cufflinks - Dynamic visualization library, wrapper for plotly, medium, example.
physt - Better histograms, talk, notebook.
matplotlib_venn - Venn diagrams, alternative.
joypy - Draw stacked density plots.
mosaic plots - Categorical variable visualization, example.
scikit-plot - ROC curves and other visualizations for ML models.
yellowbrick - Visualizations for ML models (similar to scikit-plot).
bokeh - Interactive visualization library, Examples, Examples.
lets-plot - Plotting library.
animatplot - Animate plots build on matplotlib.
plotnine - ggplot for Python.
altair - Declarative statistical visualization library.
bqplot - Plotting library for IPython/Jupyter Notebooks.
hvplot - High-level plotting library built on top of holoviews.
dtreeviz - Decision tree visualization and model interpretation.
chartify - Generate charts.
VivaGraphJS - Graph visualization (JS package).
pm - Navigatable 3D graph visualization (JS package), example.
python-ternary - Triangle plots.
falcon - Interactive visualizations for big data.

Dashboards

dash - Dashboarding solution by plot.ly. Tutorial: 1, 2, 3, 4, 5, resources
panel - Dashboarding solution.
bokeh - Dashboarding solution.
visdom - Dashboarding library by facebook.
altair example - Video.
voila - Turn Jupyter notebooks into standalone web applications.
streamlit - Dashboards.

Geographical Tools

folium - Plot geographical maps using the Leaflet.js library, jupyter plugin.
gmaps - Google Maps for Jupyter notebooks.
stadiamaps - Plot geographical maps.
datashader - Draw millions of points on a map.
sklearn - BallTree, Example.
pynndescent - Nearest neighbor descent for approximate nearest neighbors.
geocoder - Geocoding of addresses, IP addresses.
Conversion of different geo formats: talk, repo
geopandas - Tools for geographic data
Low Level Geospatial Tools (GEOS, GDAL/OGR, PROJ.4)
Vector Data (Shapely, Fiona, Pyproj)
Raster Data (Rasterio)
Plotting (Descartes, Catropy)
Predict economic indicators from Open Street Map ipynb.
PySal - Python Spatial Analysis Library.
geography - Extract countries, regions and cities from a URL or text.
cartogram - Distorted maps based on population.

Recommender Systems

Examples: 1, 2, 2-ipynb, 3.
surprise - Recommender, talk.
turicreate - Recommender.
implicit - Fast Collaborative Filtering for Implicit Feedback Datasets.
spotlight - Deep recommender models using PyTorch.
lightfm - Recommendation algorithms for both implicit and explicit feedback.
funk-svd - Fast SVD.
pywFM - Factorization.

Decision Tree Models

Intro to Decision Trees and Random Forests, Intro to Gradient Boosting
lightgbm - Gradient boosting (GBDT, GBRT, GBM or MART) framework based on decision tree algorithms, doc.
xgboost - Gradient boosting (GBDT, GBRT or GBM) library, doc, Methods for CIs: link1, link2.
catboost - Gradient boosting.
thundergbm - GBDTs and Random Forest.
h2o - Gradient boosting.
forestci - Confidence intervals for random forests.
scikit-garden - Quantile Regression.
grf - Generalized random forest.
dtreeviz - Decision tree visualization and model interpretation.
Nuance - Decision tree visualization.
rfpimp - Feature Importance for RandomForests using Permuation Importance.
Why the default feature importance for random forests is wrong: link
treeinterpreter - Interpreting scikit-learn's decision tree and random forest predictions.
bartpy - Bayesian Additive Regression Trees.
infiniteboost - Combination of RFs and GBDTs.
merf - Mixed Effects Random Forest for Clustering, video
rrcf - Robust Random Cut Forest algorithm for anomaly detection on streams.

Natural Language Processing (NLP) / Text Processing

talk-nb, nb2, talk.
Text classification Intro, Preprocessing blog post.
gensim - NLP, doc2vec, word2vec, text processing, topic modelling (LSA, LDA), Example, Coherence Model for evaluation.
Embeddings - GloVe ([1], [2]), StarSpace, wikipedia2vec, visualization.
magnitude - Vector embedding utility package.
pyldavis - Visualization for topic modelling.
spaCy - NLP.
NTLK - NLP, helpful KMeansClusterer with cosine_distance.
pytext - NLP from Facebook.
fastText - Efficient text classification and representation learning.
annoy - Approximate nearest neighbor search.
faiss - Approximate nearest neighbor search.
pysparnn - Approximate nearest neighbor search.
infomap - Cluster (word-)vectors to find topics, example.
datasketch - Probabilistic data structures for large data (MinHash, HyperLogLog).
flair - NLP Framework by Zalando.
stanfordnlp - NLP Library.
Chatistics - Turn Messenger, Hangouts, WhatsApp and Telegram chat logs into DataFrames.
textvec - Supervised text vectorization tool.

Papers

Search Engine Correlation

Biology

Sequencing

scanpy - Analyze single-cell gene expression data, tutorial.

Image-related

mahotas - Image processing (Bioinformatics), example.
imagepy - Software package for bioimage analysis.
CellProfiler - Biological image analysis.
imglyb - Viewer for large images, talk, slides.
microscopium - Unsupervised clustering of images + viewer, talk.
cytokit - Analyzing properties of cells in fluorescent microscopy datasets.

Image Processing

Talk
cv2 - OpenCV, classical algorithms: Gaussian Filter, Morphological Transformations.
scikit-image - Image processing.

Neural Networks

Tutorials & Viewer

Convolutional Neural Networks for Visual Recognition
fast.ai course - Lessons 1-7, Lessons 8-14
Tensorflow without a PhD - Neural Network course by Google.
Feature Visualization: Blog, PPT
Tensorflow Playground
Visualization of optimization algorithms, Another visualization
cutouts-explorer - Image Viewer.

Image Related

imgaug - More sophisticated image preprocessing.
Augmentor - Image augmentation library.
keras preprocessing - Preprocess images.
albumentations - Wrapper around imgaug and other libraries.
augmix - Image augmentation from Google.
kornia - Image augmentation, feature extraction and loss functions.

Lossfunction Related

SegLoss - List of loss functions for medical image segmentation.

Text Related

ktext - Utilities for pre-processing text for deep learning in Keras.
textgenrnn - Ready-to-use LSTM for text generation.
ctrl - Text generation.

Libs

keras - Neural Networks on top of tensorflow, examples.
keras-contrib - Keras community contributions.
keras-tuner - Hyperparameter tuning for Keras.
hyperas - Keras + Hyperopt: Convenient hyperparameter optimization wrapper.
elephas - Distributed Deep learning with Keras & Spark.
tflearn - Neural Networks on top of tensorflow.
tensorlayer - Neural Networks on top of tensorflow, tricks.
tensorforce - Tensorflow for applied reinforcement learning.
fastai - Neural Networks in pytorch.
pytorch-optimizer - Collection of optimizers for pytorch.
ignite - Highlevel library for pytorch.
skorch - Scikit-learn compatible neural network library that wraps pytorch, talk, slides.
autokeras - AutoML for deep learning.
PlotNeuralNet - Plot neural networks.
lucid - Neural network interpretability, Activation Maps.
tcav - Interpretability method.
AdaBound - Optimizer that trains as fast as Adam and as good as SGD, alt.
foolbox - Adversarial examples that fool neural networks.
hiddenlayer - Training metrics.
imgclsmob - Pretrained models.
netron - Visualizer for deep learning and machine learning models.
torchcv - Deep Learning in Computer Vision.

Training-related

livelossplot - Live training loss plot in Jupyter Notebook.

Object detection / Instance Segmentation

yolact - Fully convolutional model for real-time instance segmentation.
EfficientDet Pytorch, EfficientDet Keras - Scalable and Efficient Object Detection.
detectron2 - Object Detection (Mask R-CNN) by Facebook.
simpledet - Object Detection and Instance Recognition.
CenterNet - Object detection.
FCOS - Fully Convolutional One-Stage Object Detection.

Image Classification

efficientnet - Promising neural network architecture.

Applications and Snippets

CycleGAN and Pix2pix - Various image-to-image tasks.
SPADE - Semantic Image Synthesis.
Entity Embeddings of Categorical Variables, code, kaggle
Image Super-Resolution - Super-scaling using a Residual Dense Network.
Cell Segmentation - Talk, Blog Posts: 1, 2
deeplearning-models - Deep learning models.

Variational Autoencoders (VAE)

disentanglement_lib - BetaVAE, FactorVAE, BetaTCVAE, DIP-VAE.

Graph-Based Neural Networks

How to do Deep Learning on Graphs with Graph Convolutional Networks
Introduction To Graph Convolutional Networks
ogb - Open Graph Benchmark, Benchmark datasets.
networkx - Graph library.
pytorch-geometric - Various methods for deep learning on graphs.
dgl - Deep Graph Library.
graph_nets - Build graph networks in Tensorflow, by deepmind.

GPU

cuML - Run traditional tabular ML tasks on GPUs.
thundergbm - GBDTs and Random Forest.
thundersvm - Support Vector Machines.
hummingbird - Convert ML models to models that run on the GPU (by Microsoft).

Regression

Understanding SVM Regression: slides, forum, paper

pyearth - Multivariate Adaptive Regression Splines (MARS), tutorial.
pygam - Generalized Additive Models (GAMs), Explanation.
GLRM - Generalized Low Rank Models.
tweedie - Specialized distribution for zero inflated targets, Talk.

Classification

Talk, Notebook
Blog post: Probability Scoring
All classification metrics
DESlib - Dynamic classifier and ensemble selection.
human-learn - Create and tune classifier based on your rule set.

Clustering

Overview of clustering algorithms applied image data (= Deep Clustering)
Assessing the quality of a clustering (video)
hdbscan - Clustering algorithm, talk, blog.
pyclustering - All sorts of clustering algorithms.
fpc - Various methods for clustering and cluster validation (R package).
GaussianMixture - Generalized k-means clustering using a mixture of Gaussian distributions, video.
nmslib - Similarity search library and toolkit for evaluation of k-NN methods.
buckshotpp - Outlier-resistant and scalable clustering algorithm.
merf - Mixed Effects Random Forest for Clustering, video
tree-SNE - Hierarchical clustering algorithm based on t-SNE.
MiniSom - Pure Python implementation of the Self Organizing Maps.
somoclu - Self-organizing map.

Interpretable Classifiers and Regressors

skope-rules - Interpretable classifier, IF-THEN rules.
sklearn-expertsys - Interpretable classifiers, Bayesian Rule List classifier.

Multi-label classification

scikit-multilearn - Multi-label classification, talk.

Signal Processing and Filtering

Stanford Lecture Series on Fourier Transformation, Youtube, Lecture Notes.
The Scientist & Engineer's Guide to Digital Signal Processing (1999).
Kalman Filter book - Focuses on intuition using Jupyter Notebooks. Includes Baysian and various Kalman filters.
Interactive Tool for FIR and IIR filters, Examples.
filterpy - Kalman filtering and optimal estimation library.

Time Series

statsmodels - Time series analysis, seasonal decompose example, SARIMA, granger causality.
pyramid, pmdarima - Wrapper for (Auto-) ARIMA.
pyflux - Time series prediction algorithms (ARIMA, GARCH, GAS, Bayesian).
prophet - Time series prediction library.
atspy - Automated Time Series Models.
pm-prophet - Time series prediction and decomposition library.
htsprophet - Hierarchical Time Series Forecasting using Prophet.
nupic - Hierarchical Temporal Memory (HTM) for Time Series Prediction and Anomaly Detection.
tensorflow - LSTM and others, examples: link, link, link, Explain LSTM, seq2seq: 1, 2, 3, 4
tspreprocess - Preprocessing: Denoising, Compression, Resampling.
tsfresh - Time series feature engineering.
thunder - Data structures and algorithms for loading, processing, and analyzing time series data.
gatspy - General tools for Astronomical Time Series, talk.
gendis - shapelets, example.
tslearn - Time series clustering and classification, TimeSeriesKMeans, TimeSeriesKMeans.
pastas - Simulation of time series.
fastdtw - Dynamic Time Warp Distance.
fable - Time Series Forecasting (R package).
CausalImpact - Causal Impact Analysis (R package).
pydlm - Bayesian time series modeling (R package, Blog post)
PyAF - Automatic Time Series Forecasting.
luminol - Anomaly Detection and Correlation library from Linkedin.
matrixprofile-ts - Detecting patterns and anomalies, website, ppt, alternative.
stumpy - Another matrix profile library.
obspy - Seismology package. Useful classic_sta_lta function.
RobustSTL - Robust Seasonal-Trend Decomposition.
seglearn - Time Series library.
pyts - Time series transformation and classification, Imaging time series.
Turn time series into images and use Neural Nets: example, example.
sktime, sktime-dl - Toolbox for (deep) learning with time series.
adtk - Time Series Anomaly Detection.
rocket - Time Series classification using random convolutional kernels.

Time Series Evaluation

TimeSeriesSplit - Sklearn time series split.
tscv - Evaluation with gap.

Financial Data

pandas-datareader - Read stock data.
yfinance - Read stock data from Yahoo Finance.
ffn - Financial functions.
bt - Backtesting algorithms.
backtrader - Backtesting for trading strategies.
alpaca-trade-api-python - Commission-free trading through API.
eiten - Eigen portfolios, minimum variance portfolios and other algorithmic investing strategies.
tf-quant-finance - Quantitative finance tools in tensorflow, by Google.
The Quantopian Stack (some features may require signup on their platform):
pyfolio - Portfolio and risk analytics.
zipline - Algorithmic trading.
alphalens - Performance analysis of predictive stock factors.
empyrical - Financial risk metrics.
trading_calendars - Calendars for various securities exchanges.

Survival Analysis

Time-dependent Cox Model in R.
lifelines - Survival analysis, Cox PH Regression, talk, talk2.
scikit-survival - Survival analysis.
xgboost - "objective": "survival:cox" NHANES example
survivalstan - Survival analysis, intro.
convoys - Analyze time lagged conversions.
RandomSurvivalForests (R packages: randomForestSRC, ggRandomForests).
pysurvival - Survival analysis .

Outlier Detection & Anomaly Detection

sklearn - Isolation Forest and others.
pyod - Outlier Detection / Anomaly Detection.
eif - Extended Isolation Forest.
AnomalyDetection - Anomaly detection (R package).
luminol - Anomaly Detection and Correlation library from Linkedin.
Distances for comparing histograms and detecting outliers - Talk: Kolmogorov-Smirnov, Wasserstein, Energy Distance (Cramer), Kullback-Leibler divergence.
banpei - Anomaly detection library based on singular spectrum transformation.
telemanom - Detect anomalies in multivariate time series data using LSTMs.

Ranking

lightning - Large-scale linear classification, regression and ranking.

Scoring

SLIM - Scoring systems for classification, Supersparse linear integer models.

Probabilistic Modeling and Bayes

Intro, Guide
PyMC3 - Baysian modelling, intro
pomegranate - Probabilistic modelling, talk.
pmlearn - Probabilistic machine learning.
arviz - Exploratory analysis of Bayesian models.
zhusuan - Bayesian deep learning, generative models.
dowhy - Estimate causal effects.
edward - Probabilistic modeling, inference, and criticism, Mixture Density Networks (MNDs), MDN Explanation.
Pyro - Deep Universal Probabilistic Programming.
tensorflow probability - Deep learning and probabilistic modelling, talk, example.
bambi - High-level Bayesian model-building interface on top of PyMC3.
neural-tangents - Infinite Neural Networks.

Gaussian Processes

GPyOpt - Gaussian process optimization.
GPflow - Gaussian processes (Tensorflow).
gpytorch - Gaussian processes (Pytorch).

Stacking Models and Ensembles

Model Stacking Blog Post
mlxtend - EnsembleVoteClassifier, StackingRegressor, StackingCVRegressor for model stacking.
vecstack - Stacking ML models.
StackNet - Stacking ML models.
mlens - Ensemble learning.
combo - Combining ML models (stacking, ensembling).

Model Evaluation

pycm - Multi-class confusion matrix.
pandas_ml - Confusion matrix.
Plotting learning curve: link.
yellowbrick - Learning curve.

Model Explanation, Interpretability, Feature Importance

Book, Examples
shap - Explain predictions of machine learning models, talk.
treeinterpreter - Interpreting scikit-learn's decision tree and random forest predictions.
lime - Explaining the predictions of any machine learning classifier, talk, Warning (Myth 7).
lime_xgboost - Create LIMEs for XGBoost.
eli5 - Inspecting machine learning classifiers and explaining their predictions.
lofo-importance - Leave One Feature Out Importance, talk, examples: 1, 2, 3.
pybreakdown - Generate feature contribution plots.
FairML - Model explanation, feature importance.
pycebox - Individual Conditional Expectation Plot Toolbox.
pdpbox - Partial dependence plot toolbox, example.
partial_dependence - Visualize and cluster partial dependence.
skater - Unified framework to enable model interpretation.
anchor - High-Precision Model-Agnostic Explanations for classifiers.
l2x - Instancewise feature selection as methodology for model interpretation.
contrastive_explanation - Contrastive explanations.
DrWhy - Collection of tools for explainable AI.
lucid - Neural network interpretability.
xai - An eXplainability toolbox for machine learning.
innvestigate - A toolbox to investigate neural network predictions.
dalex - Explanations for ML models (R package).
interpret - Fit interpretable models, explain models (Microsoft).
causalml - Causal inference by Uber.

Automated Machine Learning

AdaNet - Automated machine learning based on tensorflow.
tpot - Automated machine learning tool, optimizes machine learning pipelines.
auto_ml - Automated machine learning for analytics & production.
autokeras - AutoML for deep learning.
nni - Toolkit for neural architecture search and hyper-parameter tuning by Microsoft.
automl-gs - Automated machine learning.
mljar - Automated machine learning.
automl_zero - Automatically discover computer programs that can solve machine learning tasks from Google.

Graph Representation Learning

Karate Club - Unsupervised learning on graphs.
Pytorch Geometric - Graph representation learning with PyTorch.
DLG - Graph representation learning with TensorFlow.

Evolutionary Algorithms & Optimization

deap - Evolutionary computation framework (Genetic Algorithm, Evolution strategies).
evol - DSL for composable evolutionary algorithms, talk.
platypus - Multiobjective optimization.
autograd - Efficiently computes derivatives of numpy code.
nevergrad - Derivation-free optimization.
gplearn - Sklearn-like interface for genetic programming.
blackbox - Optimization of expensive black-box functions.
Optometrist algorithm - paper.
DeepSwarm - Neural architecture search.

Hyperparameter Tuning

sklearn - GridSearchCV, RandomizedSearchCV.
sklearn-deap - Hyperparameter search using genetic algorithms.
hyperopt - Hyperparameter optimization.
hyperopt-sklearn - Hyperopt + sklearn.
optuna - Hyperparamter optimization, Talk.
skopt - BayesSearchCV for Hyperparameter search.
tune - Hyperparameter search with a focus on deep learning and deep reinforcement learning.
hypergraph - Global optimization methods and hyperparameter optimization.
bbopt - Black box hyperparameter optimization.
dragonfly - Scalable Bayesian optimisation.

Incremental Learning, Online Learning

sklearn - PassiveAggressiveClassifier, PassiveAggressiveRegressor.
creme-ml - Incremental learning framework, talk.
Kaggler - Online Learning algorithms.
onelearn - Online Random Forests.

Active Learning

Talk
modAL - Active learning framework.

Reinforcement Learning

YouTube, YouTube
Intro to Monte Carlo Tree Search (MCTS) - 1, 2, 3
AlphaZero methodology - 1, 2, 3, Cheat Sheet
RLLib - Library for reinforcement learning.
Horizon - Facebook RL framework.

Deployment and Lifecycle Management

Dependency Management

pipreqs - Generate a requirements.txt from import statements.
dephell - Dependency management.
poetry - Dependency management.
pyup - Dependency management.
pypi-timemachine - Install packages with pip as if you were in the past.

Data Versioning and Pipelines

dvc - Version control for large files.
hangar - Version control for tensor data.
kedro - Build data pipelines.

Data Science Related

m2cgen - Transpile trained ML models into other languages.
sklearn-porter - Transpile trained scikit-learn estimators to C, Java, JavaScript and others.
mlflow - Manage the machine learning lifecycle, including experimentation, reproducibility and deployment.
modelchimp - Experiment Tracking.
skll - Command-line utilities to make it easier to run machine learning experiments.
BentoML - Package and deploy machine learning models for serving in production.
dagster - Tool with focus on dependency graphs.
knockknock - Be notified when your training ends.
metaflow - Lifecycle Management Tool by Netflix.
cortex - Deploy machine learning models.

Math and Background

All kinds of math and statistics resources Gilbert Strang - Linear Algebra
Gilbert Strang - Matrix Methods in Data Analysis, Signal Processing, and Machine Learning

Other

daft - Render probabilistic graphical models using matplotlib.
unyt - Working with units.
scrapy - Web scraping library.
VowpalWabbit - ML Toolkit from Microsoft.
metric-learn - Metric learning.

General Python Programming

more_itertools - Extension of itertools.
funcy - Fancy and practical functional tools.
dateparser - A better date parser.
jellyfish - Approximate string matching.
coloredlogs - Colored logging output.

Resources

Distill.pub - Blog.
Machine Learning Videos
Data Science Notebooks
Recommender Systems (Microsoft)
The GAN Zoo - List of Generative Adversarial Networks
Datascience Cheatsheets

List of Books

Mat Kelceys list of cool machine learning books

Other Awesome Lists

Awesome Adversarial Machine Learning
Awesome AI Booksmarks
Awesome AI on Kubernetes
Awesome Big Data
Awesome Business Machine Learning
Awesome Causality
Awesome Community Detection
Awesome CSV
Awesome Data Science with Ruby
Awesome Dash
Awesome Decision Trees
Awesome Deep Learning
Awesome ETL
Awesome Financial Machine Learning
Awesome Fraud Detection
Awesome GAN Applications
Awesome Graph Classification
Awesome Gradient Boosting
Awesome Machine Learning
Awesome Machine Learning Interpretability
Awesome Machine Learning Operations
Awesome Monte Carlo Tree Search
Awesome Online Machine Learning
Awesome Pipeline
Awesome Python
Awesome Python Data Science
Awesome Python Data Science
Awesome Python Data Science
Awesome Pytorch
Awesome Recommender Systems
Awesome Semantic Segmentation
Awesome Sentence Embedding
Awesome Time Series
Awesome Time Series Anomaly Detection

Things I google a lot

Frequency codes for time series
Date parsing codes
Feature Calculators tsfresh

Contributing

Do you know a package that should be on this list? Did you spot a package that is no longer maintained and should be removed from this list? Then feel free to read the contribution guidelines and submit your pull request or create a new issue.

License