maxbenchrist/awesome_time_series_in_python

This curated list contains python packages for time series analysis

Using python to work with time series data

The python ecosystem contains different packages that can be used to process time series.

The following list is by no means exhaustive, feel free to edit the list (will propose a file change via PR) if you miss anything.

Machine learning, statistics, analytics

Libraries

Project Name	Description
Arrow	A sensible, human-friendly approach to creating, manipulating, formatting and converting dates, times, and timestamps
bta-lib	Technical Analysis library in pandas for backtesting algotrading and quantitative analysis
cesium	Time series platform with feature extraction aiming for non uniformly sampled signals
Darts	A library making it very easy to produce forecasts using a wide range of models, from ARIMA to deep learning. Also does ensembling, model selection and more.
ETNA	A python library for time series forecasting and analysis with temporal data structure always in mind. Includes a variety of predictive models with unified interface along with EDA and validation methods
GENDIS	Shapelet discovery by genetic algorithms
glm-sklearn	scikit-learn compatible wrapper around the GLM module in statsmodels
Featuretools	Time series feature extraction, with possible conditionality on other variables with a pandas compatible relational-database-like data container
fecon235	Computational tools for financial economics
ffn	financial function library
flint	A Time Series Library for Apache Spark
Flow Forecast	Flow Forecast is a deep learning for time series forecasting, classification, and anomaly detection framework built in PyTorch
hctsa	Matlab based feature extraction which can be controlled from python
HMMLearn	Hidden Markov Models with scikit-learn compatible API
khiva-python	A Time Series library with accelerated analytics on GPUS, it provides feature extraction and motif discovery among other functionalities.
matrixprofile-ts	Python implementation of the Matrix Profile algorithm which offers anomaly detection and pattern (or “motif”) discovery at the same time.
Nitime	Timeseries analysis for neuroscience data
Orbit	Orbit is a Python package for Bayesian time series forecasting and inference
Pandas TA	An easy to use Python 3 Pandas Extension with 130+ Technical Analysis Indicators
Pastas	Timeseries analysis for hydrological data
prophet	Time series forecasting for time series data that has multiple seasonality with linear or non-linear growth
pyDSE	ARMA models for Dynamic System Estimation
pyFTS	Fuzzy set rule-based models for time series forecasting, including multi-step, point, interval and probabilistic forecasting
PyFlux	Classical time series forecasting models
pysf	A scikit-learn compatible machine learning library for supervised/panel forecasting
pyramid	port of R's auto.arima method to Python
pytorch-forecasting	A time series forecasting library using PyTorch with various state-of-the-art network architectures.
pyts	Contains time series preprocessing, transformation as well as classification techniques
ruptures	Provides methods to find change points in time series such as shifts in the mean or scale of the signal as well as more complex changes in the probability distribution or frequency.
seglearn	Extends the scikit-learn pipeline concept to sequence data
sktime	A scikit-learn compatible library for learning with time series/panel data including time series classification/regression and (supervised/panel) forecasting
statsmodels	Contains a submodule for classical time series models and hypothesis tests
stumpy	Calculates matrix profile for time series subsequence all-pairs-similarity-search
TensorFlow-Time-Series-Examples	Time Series Prediction with tf.contrib.timeseries
tensorflow_probability.sts	Bayesian Structural Time Series model in Tensorflow Probability
timemachines	Functional interface to prophet and other packages, with Elo ratings
Traces	A library for unevenly-spaced time series analysis
ta-lib	Calculate technical indicators for financial time series (python wrapper around TA-Lib)
tsai	State-of-the-art Deep Learning with Time Series and Sequences in Pytorch / fastai
ta	Calculate technical indicators for financial time series
TIMEX	Library for creating time-series-forecasting-as-a-service platforms/websites, with a fully automated data ingestion, pre-processing, prediction and results visualization pipeline.
tsflex	A toolkit for flexible time series processing and feature extraction.
tsfresh	Extracts and filters features from time series, allowing supervised classificators and regressor to be applied to time series data
tslearn	Direct time series classifiers and regressors
tspreprocess	Preprocess time series (resampling, denoising etc.), still WIP
tsmoothie	A python library for time-series smoothing and outlier detection in a vectorized way

Examples or singular models

Project Name	Description
ES-RNN forecasting algorithm	Python implementation of the winning forecasting method of the M4 competition combining exponential smoothing with a recurrent neural network using PyTorch
Deep learning methods for time series classification	A collection of common deep learning architectures for time series classification
LSTM-Neural-Network-for-Time-Series-Prediction	LSTM based forecasting model
LSTM_tsc	An LSTM based time-series classification neural network
shapelets-python	Shapelet Classifier based on a multi layer neural network
M4 competition	Collection of statistical and machine learning forecasting methods
UCR_Time_Series_Classification_Deep_Learning_Baseline	Fully Convolutional Neural Networks for state-of-the-art time series classification
WTTE-RNN	Time to Event forecast by RNN based Weibull density estimation

Time series data container

Project name	Description
Featuretools	Time series feature extraction, with possible conditionality on other variables with a pandas compatible relational-database-like data container
pysf	A scikit-learn compatible library for supervised forecasting
xarray	Labelled, multi-dimensional data structures as long as they have a shared time index
xpandas	Labelled 1D and 2D data container for storing type-heterogeneous tabular data of any type, including time series, and encapsulates feature extraction and transformation modelling in an sklearn-compatible transformer interface, work in progress.

Data sets

Project Name	Description
awesome-public-datasets	This huge list of public datasets also has a section on time series datasets
ecmwf_models	Readers and converters for climate reanalysis data
M4 competition	Forecasting competition on 100,000 time series
pandas-datareader	Pulls financial data from different sources (e.g. yahoo, google, Quandl)
Timeseriesclassification.com	An extensive repository for time series classification datasets

Databases, frameworks

Project Name	Description
artic	High performance datastore for time series and tick data
automl_service	Fully automated time series classification pipeline, deployed as a web service
cesium	Time series platform with feature extraction aming for non uniformly sampled signals
thunder	scalable analysis of image and time series data in python based on spark
whisper	File-based time-series database format

Free courses

Project Name	Description
Time Series Forecasting	Udacity free course to learn about how to build and apply time series analysis/forecasting in business contexts

---

Discussion

We would like to trigger a homogenization of the formats which are used in the python time series community, please see the concept page

awesome_time_series_in_python's Issues

Include microprediction framework

microprediction

Add Arrow library

Arrow would also fit quite nicely here, in the general timestamps & dates handling category. From the docs:

Arrow is a Python library that offers a sensible, human-friendly approach to creating, manipulating, formatting and converting dates, times, and timestamps. It implements and updates the datetime type, plugging gaps in functionality, and provides an intelligent module API that supports many common creation scenarios. Simply put, it helps you work with dates and times with fewer imports and a lot less code.

where to find repo for multivariate time series classification or regression

great material thanks
may you clarify :
where to find repo for multivariate time series classification or regression
1
where all values are continues values
2
or even will it work for multivariate time series where values are mixture of continues and categorical values
for example 2 dimensions have continues values and 3 dimensions are categorical values

color        weight     gender  height  age  

1 black 56 m 160 34
2 white 77 f 170 54
3 yellow 87 m 167 43
4 white 55 m 198 72
5 white 88 f 176 32

Add matrixprofile-ts to list

Thanks for the great list.!

I think you should add the following: matrixprofile-ts[https://github.com/target/matrixprofile-ts]

Helps find the matrixprofile of time series, which then can be used to quickly find other information (like motifs).

Add traces library

First of all, looks like a great resource - definitely some libraries I haven't seen before, thanks.

I think traces would fit very well here. From the docs:

A Python library for unevenly-spaced time series analysis.

Updated time series software list

Hi, I took the liberty to update your list of time series related software and copy it into our wiki here, hope that's okay, of course feel free to include any of my changes into your list (I'm happy to work on this list together, but PRs seemed a little too complicated for that).

Recommend!!! HyperTS: A Full-Pipeline Automated Time Series Analysis Toolkit.

Hi,
Recommend an automatic time series toolkit HyperTS.
Github： https://github.com/DataCanvasIO/HyperTS

Easy-to-use, powerful, unified full pipeline automated time series toolkit. Supports forecasting, classification and regression.

HyperTS is a Python package that provides an end-to-end time series (TS) analysis toolkit. It covers complete and flexible AutoML workflows for TS, including data clearning, preprocessing, feature engineering, model selection, hyperparamter optimization, result evaluation, and visualization.

Multi-mode drive, light-heavy combination is the highlighted features of HyperTS. Therefore, statistical models (STATS), deep learning (DL), and neural architecture search (NAS) can be switched arbitrarily to get a powerful TS estimator.

pyFTS: Fuzzy Time Series for Python

Hi,

I would humbly like to ask you to include my library pyFTS on your awesome list. pyFTS is a free and open-source library for Fuzzy Time Series in Python programming language. Here follow some materials about it:

• Digital Object Identifier: https://doi.org/10.5281/zenodo.597359
• Official site and documentation: https://pyfts.github.io/pyFTS/
• Code repository and issue tracking: https://github.com/PYFTS/pyFTS
• PhD Thesis on Fuzzy Time Series: https://doi.org/10.5281/zenodo.3374641
• A short tutorial on Fuzzy Time Series:
  Part I: Introduction to the Fuzzy Logic, Fuzzy Time Series and the pyFTS library (https://towardsdatascience.com/a-short-tutorial-on-fuzzy-time-series-dcc6d4eb1b15)

Part II: High order, weighted and multivariate methods and a case study of solar energy forecasting. (https://towardsdatascience.com/a-short-tutorial-on-fuzzy-time-series-part-ii-with-an-case-study-on-solar-energy-bda362ecca6d )

Part III: Interval and probabilistic forecasting, non-stationary time series, concept drifts and time variant models. (https://towardsdatascience.com/a-short-tutorial-on-fuzzy-time-series-part-iii-69445dff83fb)

More example codes: https://github.com/PYFTS/notebooks

Please do not hesitate to get in touch and ask any questions. Thanks in advance.

FB Prophet against others

Sorry to ask a question using issue.
You have a tremendous article on time series analysis. Please, may you suggest a tool/python library preferably that works well for predicting a number of cars on the street at hourly level.
Right now I am using FbProphet, the predictions are off by 20% (MAPE). Trying to predict one week into the future.

The number of time series formats is too damn high!

I started a file at

https://github.com/MaxBenChrist/awesome_time_series_in_python/blob/master/standardize_time_series_formats.md

to trigger a discussion regarding the time series formats for python packages to analyze time series data.

This issue should be used to discuss the file.

Find generators for synthetic time series data sets

Thanks a lot for this awesome repo ;)
I only found a few data sets that include time series data.
Do you know a way to artificially generate time series data?
That'd be incredibly helpful.
Thanks a lot!
Willi

Include tsmoothie

I recommend you insert tsmoothie: A python library for time-series smoothing and outlier detection in a vectorized way

Link Suggestion

Thanks for organizing the time series packages in python. Here is one that is fascinating to me regarding time to event prediction. I'll let you be the judge if this qualifies for your list.

https://github.com/ragulpr/wtte-rnn/

Include darts?

The darts package is interesting. https://github.com/unit8co/darts

pyramid should be added?

https://github.com/tgsmith61591/pyramid