https://github.com/johnh2o2/cuvarbase

Python library for fast time-series analysis on CUDA GPUs
https://github.com/johnh2o2/cuvarbase

cuda fourier-methods gpu gpu-computing lomb-scargle-periodogram nfft python python-3 time-series

Last synced: about 2 months ago
JSON representation

Python library for fast time-series analysis on CUDA GPUs

• Host: GitHub
• URL: https://github.com/johnh2o2/cuvarbase
• Owner: johnh2o2
• License: gpl-3.0
• Created: 2017-06-22T14:23:18.000Z (over 7 years ago)
• Default Branch: master
• Last Pushed: 2023-10-23T01:12:48.000Z (about 1 year ago)
• Last Synced: 2024-08-09T12:27:38.207Z (5 months ago)
• Topics: cuda, fourier-methods, gpu, gpu-computing, lomb-scargle-periodogram, nfft, python, python-3, time-series
• Language: Jupyter Notebook
• Homepage: https://johnh2o2.github.io/cuvarbase/index.html
• Size: 49.9 MB
• Stars: 24
• Watchers: 7
• Forks: 4
• Open Issues: 6
• Metadata Files:
  • Readme: README.rst
  • Changelog: CHANGELOG.rst
  • License: LICENSE.txt

Awesome Lists containing this project

README

        
cuvarbase

=========

.. image:: https://badge.fury.io/py/cuvarbase.svg

    :target: https://badge.fury.io/py/cuvarbase

John Hoffman

(c) 2017

``cuvarbase`` is a Python library that uses `PyCUDA `_ to implement several time series tools used in astronomy on GPUs.

See the `documentation `_.

This project is under active development, and currently includes implementations of

- Generalized `Lomb Scargle `_ periodogram

- Box-least squares (`BLS `_ )

- Non-equispaced fast Fourier transform (adjoint operation) (`NFFT paper `_)

- Conditional entropy period finder (`CE `_)

- Phase dispersion minimization (`PDM2 `_)

	- Currently operational but minimal unit testing or documentation (yet)

Hopefully future developments will have

- (Weighted) wavelet transforms

- Spectrograms (for PDM and GLS)

- Multiharmonic extensions for GLS

Dependencies

------------

- `PyCUDA `_ **<-essential**

- `scikit cuda `_ **<-also essential**

	- used for access to the CUDA FFT runtime library

- `matplotlib `_ (for plotting utilities)

- `nfft `_ (for unit testing)

- `astropy `_ (for unit testing)

Using multiple GPUs

-------------------

If you have more than one GPU, you can choose which one to

use in a given script by setting the ``CUDA_DEVICE`` environment

variable:

.. code:: sh

    CUDA_DEVICE=1 python script.py

If anyone is interested in implementing multi-device load-balancing

solution, they are encouraged to do so! At some point this may

become important, but for the time being manually splitting up the

jobs to different GPU's will have to suffice.