TE
TechEcho
Home24h TopNewestBestAskShowJobs
GitHubTwitter
Home

TechEcho

A tech news platform built with Next.js, providing global tech news and discussions.

GitHubTwitter

Home

HomeNewestBestAskShowJobs

Resources

HackerNews APIOriginal HackerNewsNext.js

© 2025 TechEcho. All rights reserved.

Darts: A Python library for easy manipulation and forecasting of time series

239 pointsby grzraczalmost 4 years ago

10 comments

hrznalmost 4 years ago
Hi! I&#x27;m one core developer (and creator) of the library. Thanks for all the comments. I just wanted to highlight a couple of things that we think are quite cool about Darts:<p>* It makes using all sorts of forecasting models (from ARIMA to deep learning) easy, using fit() and predict(), similar to scikit-learn.<p>* It&#x27;s easy to fit deep learning and other ML-based models on multiple time series, potentially on big datasets too. The time series can be multivariate.<p>* Darts is not only wrapping existing models. We also have our own implementations, for instance of TCN (Temporal Convolutional Networks), or adaptations N-BEATS (which we extended to handle multivariate series), DeepAR and others.<p>* Darts makes it very easy to include past and&#x2F;or future covariates as inputs for the predictions.<p>* Some models offer probabilistic forecasts; sometimes with the possibility to configure your favourite likelihood function (e.g. Gaussian for continuous values or Poisson for discrete values).<p>* Everything uses the &quot;TimeSeries&quot; class, which makes the API consistent across tools and models, and make it harder to make mistakes. For instance it&#x27;s easy to consume the output of one model by another model, and all models can be backtested the same way.
评论 #28156353 未加载
评论 #28157915 未加载
评论 #28156560 未加载
评论 #28163717 未加载
XoS-490almost 4 years ago
Hey guys,<p>If you are interested in a unified interface for time series learning tasks, have a look on sktime:<p><a href="https:&#x2F;&#x2F;github.com&#x2F;alan-turing-institute&#x2F;sktime" rel="nofollow">https:&#x2F;&#x2F;github.com&#x2F;alan-turing-institute&#x2F;sktime</a><p><a href="https:&#x2F;&#x2F;www.sktime.org&#x2F;en&#x2F;stable&#x2F;" rel="nofollow">https:&#x2F;&#x2F;www.sktime.org&#x2F;en&#x2F;stable&#x2F;</a><p>It provides specialized time series algorithms and scikit-learn compatible tools to build, tune and validate time series models for multiple learning problems.<p>We have been working hard on this open-source project for a long time and would be glad to hear your opinion.
评论 #28163367 未加载
chudifiedalmost 4 years ago
Darts is mostly a wrapper for a bunch of other timeseries forecasting libraries, and provides a single interface to work with them. It&#x27;s not really a Facebook alternative. Facebook&#x27;s Prophet library is one of the forecasting libraries used by Darts.
评论 #28155624 未加载
评论 #28155416 未加载
peter_retiefalmost 4 years ago
I am writing code for a time series NN at the moment so am keen to check out darts<p>conda install -c conda-forge -c pytorch pip prophet pytorch cpuonly<p>pip install darts<p>Why not install darts with conda as well?<p>I am getting some errors runnning the on a debian system install<p>when running the example script I get this: Importing plotly failed. Interactive plots will not work. &#x2F;home&#x2F;peter&#x2F;anaconda3&#x2F;envs&#x2F;darts&#x2F;lib&#x2F;python3.7&#x2F;site-packages&#x2F;statsmodels&#x2F;tsa&#x2F;holtwinters&#x2F;model.py:429: FutureWarning: After 0.13 initialization must be handled at model creation FutureWarning,<p>I guess its a library compatabilty but feedback is important.
评论 #28173880 未加载
bartkappenburgalmost 4 years ago
On a side note: I love the example with the airtravel time series (# of passengers). It fits almost perfectly but I love to see how it holds up with data from around March 2020 until now ;-).<p>Disclaimer: studied econometrics so I will try do this on my own :D
bothra90almost 4 years ago
How does this compare with <a href="https:&#x2F;&#x2F;github.com&#x2F;linkedin&#x2F;greykite" rel="nofollow">https:&#x2F;&#x2F;github.com&#x2F;linkedin&#x2F;greykite</a>?
评论 #28161005 未加载
kleibaalmost 4 years ago
Can&#x27;t wait for the Dart port.
评论 #28180892 未加载
jstx1almost 4 years ago
&gt; Darts: Non-Facebook alternative for timeseries forecasting<p>The title of the post seems very editorialised.<p>First of all, being non-Facebook is hardly meaningful when we talk about open source tools. Secondly, the project doesn&#x27;t advertise itself as being non-Facebook, the poster has added this. And lastly, it&#x27;s false - from the prerequisites in the readme:<p>&gt; Some of the models depend on prophet and torch
评论 #28159441 未加载
评论 #28155400 未加载
bernardvalmost 4 years ago
What has Facebook got to do with time-series forecasting?? Prophet?? Real time series work means rolling your own tools and understanding exactly how they function ;-). In C preferably, while flogging yourself at the same time. Just kidding.. Sort of.
olavalmost 4 years ago
I opened this page because of the friendly &quot;Non-Facebook alternative...&quot;. In my view, we should ignore software by companies that actively seek to destroy our societies.