Skip to content

Release 0.10.0
Compare
Choose a tag to compare
@dizcza dizcza released this 04 Mar 18:37
· 151 commits to master since this release
v0.10.0
165276f
This commit was created on GitHub.com and signed with GitHub’s verified signature. The key has expired.
GPG key ID: 4AEE18F83AFDEB23
Expired
Learn about vigilant mode.

Documentation

The documentation is revised and restructured by categories (#386) to simplify navigation on readthedocs and improve user experience. All citations used in Elephant are stored in a single BibTex file.

Optimizations

CUDA and OpenCL support

Analysis of Sequences of Synchronous EvenTs has become the first module in Elephant that supports CUDA and OpenCL (#351, #404, #399). Whether you have an Nvidia GPU or just run the analysis on a laptop with a built-in Intel graphics card, the speed-up is X100 and X1000 compared to a single CPU core. The computations are optimized to a degree that you can analyse and look for spike patterns in real data in several minutes of compute time on a laptop. The installation instructions are described in the install section.

Other optimizations

  • Surrogates: sped up bin shuffling (#400) and reimplemented the continuous time version (#397)
  • Improved memory efficiency of creating a BinnedSpikeTrain (#395)

New functionality and features

  • Synchrofact detection (#322) is a method to detect highly synchronous spikes (at the level of sampling rate precision with an option to extend this to jittered synchrony) and annotate or optionally remove them.
  • Added phase_locking_value, mean_phase_vector, and phase_difference functions (https://github.com/NeuralEnsemble/elephant/pull/385/files)
  • BinnedSpikeTrain:
    • added to_spike_trains and time_slice functions (#390). Now you can slice a binned spike train as bst[:, i:j] or bst.time_slice(t_start, t_stop). Also, with to_spike_trains function, you can generate a realization of spike trains that maps to the same BinnedSpikeTrain object when binned.
    • optional CSC format (#402)
    • the copy parameter (False by default) in the binarize function makes a shallow copy, if set to True, of the output BinnedSpikeTrain object (#402)
  • Granger causality tutorial notebook (#393)
  • Unitary Event Analysis support multiple pattern hashes (#387)

Bug fixes

  • Account for unidirectional spiketrain->segment links in synchrofact deletion (#398)
  • Joint-ISI dithering: fixed a bug regarding first ISI bin (#396)
  • Fix LvR values from being off when units are in seconds (#389)
Assets 2
Loading