Skip to content

Commit 2888352

Browse files
authored
Update README (#104)
* Remove conda badge * Fix linting issue * Update installation instructions * Remove empty benchmarks section * Fix spacing * Fix wording * Update installation requirements * Update links and citation
1 parent 0db5702 commit 2888352

File tree

1 file changed

+21
-40
lines changed

1 file changed

+21
-40
lines changed

README.md

Lines changed: 21 additions & 40 deletions
Original file line numberDiff line numberDiff line change
@@ -8,7 +8,6 @@
88
[![Quality](https://github.com/Rose-STL-Lab/torchTS/workflows/Quality/badge.svg)](https://github.com/Rose-STL-Lab/torchTS/actions?query=workflow%3AQuality)
99
[![Docs](https://github.com/Rose-STL-Lab/torchTS/workflows/Docs/badge.svg)](https://github.com/Rose-STL-Lab/torchTS/actions?query=workflow%3ADocs)
1010
[![Codecov](https://img.shields.io/codecov/c/github/Rose-STL-Lab/torchTS?label=Coverage&logo=codecov)](https://app.codecov.io/gh/Rose-STL-Lab/torchTS)
11-
[![Conda](https://img.shields.io/conda/v/pytorch/torchts?label=Conda&logo=anaconda)](https://anaconda.org/pytorch/torchts)
1211
[![PyPI](https://img.shields.io/pypi/v/torchts?label=PyPI&logo=python)](https://pypi.org/project/torchts)
1312
[![License](https://img.shields.io/github/license/Rose-STL-Lab/torchTS?label=License)](LICENSE)
1413

@@ -18,35 +17,34 @@ TorchTS is a PyTorch-based library for time series data.
1817

1918
#### Why Time Series?
2019

21-
Time series data modeling has broad significance in public health, finance and engineering. Traditional time series methods from statistics often rely on strong modeling assumptions, or are computationally expensive. Given the rise of large-scale sensing data and significant advances in deep learning, the goal of the project is to develop an efficient and user-friendly deep learning library that would benefit the entire research community and beyond.
20+
Time series data modeling has broad significance in public health, finance and engineering. Traditional time series methods from statistics often rely on strong modeling assumptions, or are computationally expensive. Given the rise of large-scale sensing data and significant advances in deep learning, the goal of the project is to develop an efficient and user-friendly deep learning library that would benefit the entire research community and beyond.
2221

2322
#### Why TorchTS?
2423

25-
Existing time series analysis libraries include [statsmodels](https://www.statsmodels.org/stable/index.html), [sktime](https://github.com/alan-turing-institute/sktime). However, these libraries only include traditional statistics tools such as ARMA or ARIMA, which do not have the state-of-the-art forecasting tools based on deep learning. [GluonTS](https://ts.gluon.ai/) is an open-source time series library developed by Amazon AWS, but is based on MXNet. [Pyro](https://pyro.ai/) is a probabilistic programming framework based on PyTorch, but is not focused on time series forecasting.
26-
27-
#### Benchmark
24+
Existing time series analysis libraries include [statsmodels](https://www.statsmodels.org/stable/index.html) and [sktime](https://github.com/alan-turing-institute/sktime). However, these libraries only include traditional statistics tools such as ARMA or ARIMA, which do not have the state-of-the-art forecasting tools based on deep learning. [GluonTS](https://ts.gluon.ai/) is an open-source time series library developed by Amazon AWS, but is based on MXNet. [Pyro](https://pyro.ai/) is a probabilistic programming framework based on PyTorch, but is not focused on time series forecasting.
2825

2926
## Installation
3027

3128
### Installation Requirements
3229

33-
- Python >= 3.7
34-
- PyTorch >= 1.7
35-
- scipy
30+
TorchTS supports Python 3.7+ and has the following dependencies:
31+
32+
- [PyTorch](https://pytorch.org/)
33+
- [PyTorch Lightning](https://pytorchlightning.ai/)
34+
- [SciPy](https://www.scipy.org/)
3635

3736
### Installing the latest release
3837

39-
The latest release of TorchTS is easily installed either via
40-
[Anaconda](https://www.anaconda.com/distribution/#download-section) (recommended):
38+
The latest release of TorchTS is easily installed either via `pip`:
4139

4240
```bash
43-
conda install torchts -c pytorch
41+
pip install torchts
4442
```
4543

46-
or via `pip`:
44+
or via [conda](https://docs.conda.io/projects/conda/) from the [conda-forge](https://conda-forge.org/) channel (coming soon):
4745

4846
```bash
49-
pip install torchts
47+
conda install torchts -c conda-forge
5048
```
5149

5250
You can customize your PyTorch installation (i.e. CUDA version, CPU only option)
@@ -63,48 +61,31 @@ by following the [PyTorch installation instructions](https://pytorch.org/get-sta
6361

6462
## Getting Started
6563

66-
Here's a quick run down of the main components of a deep learning forecaster.
67-
For more details see our [Documentation](https://rose-stl-lab.github.io/torchTS/) and the
68-
[Tutorials](https://torchts.ai/tutorials).
69-
70-
1. Multi-step forecasting
71-
72-
```python
73-
74-
```
75-
76-
2. Spatiotemporal forecasting
77-
```python
78-
79-
```
80-
81-
3. Hybrid forecasting with PDEs
82-
83-
```python
84-
85-
```
64+
Check out our [documentation](https://rose-stl-lab.github.io/torchTS/) and
65+
[tutorials](https://rose-stl-lab.github.io/torchTS/tutorials) (coming soon).
8666

8767
## Citing TorchTS
8868

89-
If you use TorchTS, please cite the following paper:
90-
> [TBD. TorchTS: A Framework for Efficient Time Series Modeling.](TBD)
69+
If you use TorchTS, please cite the following paper (coming soon):
70+
71+
> [TorchTS: A Framework for Efficient Time Series Modeling](TBD)
9172
9273
```bibtex
9374
@inproceedings{TBD,
9475
title={{TorchTS: A Framework for Efficient Time Series Modeling}},
95-
author={},
76+
author={TBD},
9677
booktitle = {TBD},
9778
year={TBD},
98-
url = {}
79+
url = {TBD}
9980
}
10081
```
10182

102-
See [here](https://torchts.ai/docs/papers) for an incomplete selection of peer-reviewed papers that build off of TorchTS.
83+
See [here](https://rose-stl-lab.github.io/torchTS/papers) (coming soon) for a selection of peer-reviewed papers that either build off of TorchTS or were integrated into TorchTS.
10384

10485
## Contributing
10586

106-
See the [CONTRIBUTING](CONTRIBUTING.md) file for how to help out.
87+
Interested in contributing to TorchTS? Please see the [contributing guide](CONTRIBUTING.md) to learn how to help out.
10788

10889
## License
10990

110-
TorchTS is MIT licensed, as found in the [LICENSE](LICENSE) file.
91+
TorchTS is [MIT licensed](LICENSE).

0 commit comments

Comments
 (0)