{ "id": 286292, "title": "Useful libraries", "url": "/competitions/g-research-crypto-forecasting/discussion/286292", "author_name": "", "post_date": "2021-11-08T16:50:41.815255300Z", "votes": 14, "comment_count": 8, "views": 0, "content": "

Hello everyone,

\n

I'm doing some research on time series for trading and found these libraries.
\nCan you share your thoughts on them?
\nDo you think they are appropriate for this project?

\n\n

Also, would you consider using Databricks, Dask, or Rapids here? I know Databricks has some financial applications.

\n

——— Update
\nGluon Time Series (GluonTS) is the Gluon toolkit for probabilistic time series modeling, focusing on deep learning-based models.
\nExample by Stephan Rabanser on the M5 Forecasting - Accuracy challenge

", "messages": [ { "id": "1575801", "postDate": "11/08/2021 16:50:41", "content": "

Hello everyone,

\n

I'm doing some research on time series for trading and found these libraries.
\nCan you share your thoughts on them?
\nDo you think they are appropriate for this project?

\n\n

Also, would you consider using Databricks, Dask, or Rapids here? I know Databricks has some financial applications.

\n

——— Update
\nGluon Time Series (GluonTS) is the Gluon toolkit for probabilistic time series modeling, focusing on deep learning-based models.
\nExample by Stephan Rabanser on the M5 Forecasting - Accuracy challenge

", "rawMarkdown": "Hello everyone,\n\nI'm doing some research on time series for trading and found these libraries.\nCan you share your thoughts on them?\nDo you think they are appropriate for this project? \n\n- [Prophet](https://facebook.github.io/prophet/docs/quick_start.html): Well known, follows the sklearn model API.\n- [Darts](https://unit8co.github.io/darts/) Contains a variety of models, from classics such as ARIMA to deep neural networks. Promises to be easy to use.\n- [Nixtla](https://github.com/Nixtla) These guys are doing amazing stuff. Keep an eye on them.\n- [TA-Lib](https://mrjbq7.github.io/ta-lib/): Supported Indicators\n- [Stumpy](https://github.com/TDAmeritrade/stumpy): A Python library that can be used for a variety of time series data mining tasks\n- [Cesium-ML](https://cesium-ml.org/docs/auto_examples/plot_EEG_Example.html#sphx-glr-auto-examples-plot-eeg-example-py): Open-Source machine learning for time series analysis\n- [tsfresh](https://tsfresh.readthedocs.io/en/latest/): Automatic extraction of relevant features from time series\n- [SHAP](https://github.com/slundberg/shap): Game-theory approach to feature importance.\n- [Hctsa](https://tsfresh.readthedocs.io/en/latest/text/introduction.html)+[Pyopy](https://github.com/strawlab/pyopy): [matlab adapt] automatically extract features from time series\n- [FATS](http://isadoranun.github.io/tsfeat/): (Feature Analysis for Time Series) For feature extraction from time-series data.\n- [Featuretools](https://www.featuretools.com/): Automated feature engineering\n- [Alphalens](https://alphalens.ml4trading.io/notebooks/overview.html): Performance analysis of predictive (alpha) stock factors\n- [pyfolio](https://quantopian.github.io/pyfolio/) is a Python library for the performance and risk analysis of financial portfolios.\n- [Empyrical](http://quantopian.github.io/empyrical/): Common financial risk metrics.\n- [Tensorflow-time_series](https://www.tensorflow.org/tutorials/structured_data/time_series)\n- [Masked time series](https://www.tensorflow.org/probability/api_docs/python/tfp/sts/MaskedTimeSeries) For dealing with missing values\n\nAlso, would you consider using Databricks, Dask, or Rapids here? I know Databricks has some [financial applications](https://www.youtube.com/watch?v=4cfC26NsJGI&).\n\n——— Update\n[Gluon Time Series (GluonTS)](https://ts.gluon.ai/index.html) is the Gluon toolkit for probabilistic time series modeling, focusing on deep learning-based models.\n[Example](https://www.kaggle.com/steverab/m5-forecasting-competition-gluonts-template) by Stephan Rabanser on the M5 Forecasting - Accuracy challenge", "votes": null }, { "id": "1576774", "postDate": "11/09/2021 13:04:26", "content": "

There is also Pandas Technical Analysis, an easy-to-use library that leverages the Pandas package with more than 130 Indicators and Utility functions and more than 60 TA-Lib Candlestick Patterns.

", "rawMarkdown": "There is also [Pandas Technical Analysis](https://github.com/twopirllc/pandas-ta), an easy-to-use library that leverages the Pandas package with more than 130 Indicators and Utility functions and more than 60 TA-Lib Candlestick Patterns.", "votes": null }, { "id": "1577157", "postDate": "11/09/2021 20:19:19", "content": "

For those who find difficulties installing TA-Lib on a Kaggle notebook, use this code:

\n
%cd /kaggle/working\n%rm -rf temp\n!mkdir temp\n%cd ./temp\n!wget http://prdownloads.sourceforge.net/ta-lib/ta-lib-0.4.0-src.tar.gz\n!tar -xzvf ta-lib-0.4.0-src.tar.gz\n%cd ./ta-lib\n!./configure --prefix=/usr\n!make\n!make install\n!pip install Ta-Lib\n%cd /kaggle/working\n!rm -rf temp\n
", "rawMarkdown": "For those who find difficulties installing `TA-Lib` on a Kaggle notebook, use this code:\n```\n%cd /kaggle/working\n%rm -rf temp\n!mkdir temp\n%cd ./temp\n!wget http://prdownloads.sourceforge.net/ta-lib/ta-lib-0.4.0-src.tar.gz\n!tar -xzvf ta-lib-0.4.0-src.tar.gz\n%cd ./ta-lib\n!./configure --prefix=/usr\n!make\n!make install\n!pip install Ta-Lib\n%cd /kaggle/working\n!rm -rf temp\n```", "votes": null }, { "id": "1578302", "postDate": "11/10/2021 20:53:12", "content": "

To commit our notebooks we have to disable internet so how can we install external libraries during commit ?
\nThere is any way to make the installation permanent ?
\nThx

", "rawMarkdown": "To commit our notebooks we have to disable internet so how can we install external libraries during commit ?\nThere is any way to make the installation permanent ?\nThx", "votes": null }, { "id": "1578370", "postDate": "11/11/2021 00:28:08", "content": "

This might work for you
\nhttps://www.kaggle.com/samuelepino/pip-installing-packages-with-no-internet

", "rawMarkdown": "This might work for you\nhttps://www.kaggle.com/samuelepino/pip-installing-packages-with-no-internet", "votes": null }, { "id": "1578644", "postDate": "11/11/2021 08:32:22", "content": "

Thanks , but it does not work , when I restart the session or commit, the content of directory working is lost …

", "rawMarkdown": "Thanks , but it does not work , when I restart the session or commit, the content of directory working is lost ...", "votes": null }, { "id": "1578671", "postDate": "11/11/2021 08:59:16", "content": "

It seems to be working … Many tahnks .. here the steps

\n
    \n
  1. Upload you library as dataset as libs(or choose your dataset name name)
    2. \n
  2. ! pip install libname file:../input/libs/dist/file.tar
    3. \n
\n

Many thanks for your hint 👍

", "rawMarkdown": "It seems to be working ... Many tahnks .. here the steps\n1. Upload you library as dataset as libs(or choose your dataset name name)\n2. ! pip install libname file:../input/libs/dist/file.tar\n\nMany thanks for your hint 👍", "votes": null }, { "id": "1578673", "postDate": "11/11/2021 09:04:45", "content": "

one of the best threads related to the Competition. thanks for your efforts.

", "rawMarkdown": "one of the best threads related to the Competition. thanks for your efforts.", "votes": null }, { "id": "1579153", "postDate": "11/11/2021 15:34:44", "content": "

I am glad to read that.

", "rawMarkdown": "I am glad to read that.", "votes": null } ], "comments": [ { "id": 1576774, "author_name": "abdelghanibelgaid", "author_url": "", "post_date": "11/09/2021 13:04:26", "content": "

There is also Pandas Technical Analysis, an easy-to-use library that leverages the Pandas package with more than 130 Indicators and Utility functions and more than 60 TA-Lib Candlestick Patterns.

", "votes": null, "replies": [] }, { "id": 1577157, "author_name": "abdelghanibelgaid", "author_url": "", "post_date": "11/09/2021 20:19:19", "content": "

For those who find difficulties installing TA-Lib on a Kaggle notebook, use this code:

\n
%cd /kaggle/working\n%rm -rf temp\n!mkdir temp\n%cd ./temp\n!wget http://prdownloads.sourceforge.net/ta-lib/ta-lib-0.4.0-src.tar.gz\n!tar -xzvf ta-lib-0.4.0-src.tar.gz\n%cd ./ta-lib\n!./configure --prefix=/usr\n!make\n!make install\n!pip install Ta-Lib\n%cd /kaggle/working\n!rm -rf temp\n
", "votes": null, "replies": [] }, { "id": 1578302, "author_name": "gianpaolobulleddu", "author_url": "", "post_date": "11/10/2021 20:53:12", "content": "

To commit our notebooks we have to disable internet so how can we install external libraries during commit ?
\nThere is any way to make the installation permanent ?
\nThx

", "votes": null, "replies": [ { "id": 1578370, "author_name": "echo9k", "author_url": "", "post_date": "11/11/2021 00:28:08", "content": "

This might work for you
\nhttps://www.kaggle.com/samuelepino/pip-installing-packages-with-no-internet

", "votes": null, "replies": [] }, { "id": 1578644, "author_name": "gianpaolobulleddu", "author_url": "", "post_date": "11/11/2021 08:32:22", "content": "

Thanks , but it does not work , when I restart the session or commit, the content of directory working is lost …

", "votes": null, "replies": [] }, { "id": 1578671, "author_name": "gianpaolobulleddu", "author_url": "", "post_date": "11/11/2021 08:59:16", "content": "

It seems to be working … Many tahnks .. here the steps

\n
    \n
  1. Upload you library as dataset as libs(or choose your dataset name name)
    2. \n
  2. ! pip install libname file:../input/libs/dist/file.tar
    3. \n
\n

Many thanks for your hint 👍

", "votes": null, "replies": [] }, { "id": 1579153, "author_name": "echo9k", "author_url": "", "post_date": "11/11/2021 15:34:44", "content": "

I am glad to read that.

", "votes": null, "replies": [] } ] }, { "id": 1578673, "author_name": "yasinnaal", "author_url": "", "post_date": "11/11/2021 09:04:45", "content": "

one of the best threads related to the Competition. thanks for your efforts.

", "votes": null, "replies": [] } ], "raw_markdown_by_id": { "1575801": "Hello everyone,\n\nI'm doing some research on time series for trading and found these libraries.\nCan you share your thoughts on them?\nDo you think they are appropriate for this project? \n\n- [Prophet](https://facebook.github.io/prophet/docs/quick_start.html): Well known, follows the sklearn model API.\n- [Darts](https://unit8co.github.io/darts/) Contains a variety of models, from classics such as ARIMA to deep neural networks. Promises to be easy to use.\n- [Nixtla](https://github.com/Nixtla) These guys are doing amazing stuff. Keep an eye on them.\n- [TA-Lib](https://mrjbq7.github.io/ta-lib/): Supported Indicators\n- [Stumpy](https://github.com/TDAmeritrade/stumpy): A Python library that can be used for a variety of time series data mining tasks\n- [Cesium-ML](https://cesium-ml.org/docs/auto_examples/plot_EEG_Example.html#sphx-glr-auto-examples-plot-eeg-example-py): Open-Source machine learning for time series analysis\n- [tsfresh](https://tsfresh.readthedocs.io/en/latest/): Automatic extraction of relevant features from time series\n- [SHAP](https://github.com/slundberg/shap): Game-theory approach to feature importance.\n- [Hctsa](https://tsfresh.readthedocs.io/en/latest/text/introduction.html)+[Pyopy](https://github.com/strawlab/pyopy): [matlab adapt] automatically extract features from time series\n- [FATS](http://isadoranun.github.io/tsfeat/): (Feature Analysis for Time Series) For feature extraction from time-series data.\n- [Featuretools](https://www.featuretools.com/): Automated feature engineering\n- [Alphalens](https://alphalens.ml4trading.io/notebooks/overview.html): Performance analysis of predictive (alpha) stock factors\n- [pyfolio](https://quantopian.github.io/pyfolio/) is a Python library for the performance and risk analysis of financial portfolios.\n- [Empyrical](http://quantopian.github.io/empyrical/): Common financial risk metrics.\n- [Tensorflow-time_series](https://www.tensorflow.org/tutorials/structured_data/time_series)\n- [Masked time series](https://www.tensorflow.org/probability/api_docs/python/tfp/sts/MaskedTimeSeries) For dealing with missing values\n\nAlso, would you consider using Databricks, Dask, or Rapids here? I know Databricks has some [financial applications](https://www.youtube.com/watch?v=4cfC26NsJGI&).\n\n——— Update\n[Gluon Time Series (GluonTS)](https://ts.gluon.ai/index.html) is the Gluon toolkit for probabilistic time series modeling, focusing on deep learning-based models.\n[Example](https://www.kaggle.com/steverab/m5-forecasting-competition-gluonts-template) by Stephan Rabanser on the M5 Forecasting - Accuracy challenge", "1576774": "There is also [Pandas Technical Analysis](https://github.com/twopirllc/pandas-ta), an easy-to-use library that leverages the Pandas package with more than 130 Indicators and Utility functions and more than 60 TA-Lib Candlestick Patterns.", "1577157": "For those who find difficulties installing `TA-Lib` on a Kaggle notebook, use this code:\n```\n%cd /kaggle/working\n%rm -rf temp\n!mkdir temp\n%cd ./temp\n!wget http://prdownloads.sourceforge.net/ta-lib/ta-lib-0.4.0-src.tar.gz\n!tar -xzvf ta-lib-0.4.0-src.tar.gz\n%cd ./ta-lib\n!./configure --prefix=/usr\n!make\n!make install\n!pip install Ta-Lib\n%cd /kaggle/working\n!rm -rf temp\n```", "1578302": "To commit our notebooks we have to disable internet so how can we install external libraries during commit ?\nThere is any way to make the installation permanent ?\nThx", "1578370": "This might work for you\nhttps://www.kaggle.com/samuelepino/pip-installing-packages-with-no-internet", "1578644": "Thanks , but it does not work , when I restart the session or commit, the content of directory working is lost ...", "1578671": "It seems to be working ... Many tahnks .. here the steps\n1. Upload you library as dataset as libs(or choose your dataset name name)\n2. ! pip install libname file:../input/libs/dist/file.tar\n\nMany thanks for your hint 👍", "1578673": "one of the best threads related to the Competition. thanks for your efforts.", "1579153": "I am glad to read that." }, "source": "meta" }