1.6.1

• Safer parsing of channel names in the legacy stream viewer, ensuring that incomplete metadata in the StreamInfo should be correctly parsed in the correct number of channels.

Full Changelog: 1.6.0...1.6.1

1.6.0

What's Changed

• Improve coverage and improve disconnection of epochs by @mscheltienne in #317
• Fix sample and testing datasets path by @mscheltienne in #319
• Drop isort and pydocstyle in favor of ruff and pin to 0.6.0 and above by @mscheltienne in #321
• Get rid of qt5 entirely in optional requirements by @mscheltienne in #326
• add mailmap and enforce sorting of it by @drammock in #327
• Draft JOSS paper and add PDF generation gh-action by @mscheltienne in #316
• Fix negative start index when looping around dataset by @toni-neurosc in #330
• Add contributing guide by @mscheltienne in #332

New Contributors

• @drammock made their first contribution in #327
• @toni-neurosc made their first contribution in #330

Full Changelog: 1.5.0...1.6.0

1.5.0

What's Changed

• Use property n_new_samples in detector examlpe by @mscheltienne in #288
• Add EpochsStream for epoching real-time stream on-the-fly by @mscheltienne in #258
• Add tutorial about epochs and expose base stream class by @mscheltienne in #302
• CI: Fix retry action by @larsoner in #306
• Make unit tests more robust and add tutorial for Player with annotations by @mscheltienne in #305
• Improve peak detection example using chunk size 1 by @mscheltienne in #308
• Add documentation about processing steps by @mscheltienne in #309
• Add real-time decoding example by @mscheltienne in #312

Full Changelog: 1.4.0...1.5.0

1.4.0

What's Changed

• Improve warning handling by @mscheltienne in #243
• Fix CLI entry-point for Player by @mscheltienne in #246
• Fix push operation with a player having chunk_size set to 1 by @mscheltienne in #257
• Add example for a real-time peak detector applied to cardiac R-peak by @mscheltienne in #256
• Replace single-use thread operation with a ThreadPoolExecutor by @mscheltienne in #264
• Add example of player run in child process by @mscheltienne in #267
• Fix pushing the last chunk of a finite mock stream with chunk size of 1 by @mscheltienne in #268
• Fix flaky test by @mscheltienne in #270
• Improve CIs, example and code coverage by @mscheltienne in #273
• Improve legacy viewer scaling by @mscheltienne in #275
• Add example for bandpower on rolling window by @mscheltienne in #279
• Try to fix segfault in CIs during inlet/outlet deletion/destruction by @mscheltienne in #277
• Add manual acquisition method by @mscheltienne in #282

Full Changelog: 1.3.1...1.4.0

1.3.1

• Improve handling of warnings throughout the codebase to improve the feedback received by @mscheltienne in #243

Full Changelog: 1.3.0...1.3.1

1.3.0

What's Changed

• Add IIR filters to a Stream object through the methods Stream.filter and Stream.notch_filter by @mscheltienne and @larsoner in #218 and #226
• Improve logging and level handling by @mscheltienne in #219
• Add n_repeat argument to Player by @mscheltienne in #223
• Improve documentation of performance differences between pylsl and mne.lsl by @mscheltienne in #231

Full Changelog: 1.2.0...1.3.0

1.2.0

• Implement MNE's design philosophy by returning self in methods modifying a StreamLSL -> enables method chaining stream.pick("eeg").set_eeg_reference("average")
• Add argument annotations to PlayerLSL to stream annotations on a separate irregularly sampled stream
• Add support for loaded Raw objects as input to a PlayerLSL
• Improve deletion or low-level resources
• Improve discovery and download of existing liblsl on the system
• Fix reading of yRange through settings in legacy StreamViewer

Full Changelog: 1.1.1...1.2.0

1.1.1

• Fix parsing of incomplete StreamInfo into an mne.Info object by @mscheltienne in #179
• Remove pytest related files from the distribution by @mscheltienne in #180

Full Changelog: 1.1.0...1.1.1

1.1.0

What's Changed

• Remove legacy setup.py by @mscheltienne in #167
• Improve documentation workflow by @mscheltienne in #170
• Add support for 'timestamp' array in a push operation by @mscheltienne in #172
• Add local sphinx-remove-toctrees extension by @mscheltienne in #173
• Improve type-hints by @mscheltienne in #175
• MAINT: Make tests less flaky by @larsoner in #168
• Improve test coverage on load-liblsl and on datasets by @mscheltienne in #176
• Improve test coverage, structure and checks for strean connection in properties by @mscheltienne in #177
• Match argument order between BasePlayer and PlayerLSL by @mscheltienne in #178

New Contributors

• @larsoner made their first contribution in #168

Full Changelog: 1.0.0...1.1.0

Please report encountered issues on the tracker.

1.0.0

Initial release of mne-lsl for Python 3.9 to 3.12. MNE-LSL integrates tightly LSL streams with MNE-Python. It replaces the LSLClient in mne_realtime with 2 main objects:

• mne_lsl.stream.StreamLSL: connects to an LSL stream and acts as a Raw object which underlying data array is a ringbuffer continuously updated.
• mne_lsl.player.PlayerLSL: create a fake LSL stream from any MNE-readable file.

player = mne_lsl.player.PlayerLSL(...)
player.start()
stream = mne_lsl.stream.StreamLSL(...)
stream.connect()
stream.pick("eeg")
stream.set_eeg_reference("average")
data, ts = stream.get_data()
stream.disconnect()
player.stop()

Additionally, the low-level mne_lsl.lsl module re-implements a faster version of pylsl with convenience methods to set and retrieve metadata, for instance from an mne.Info object.

sinfo = mne_lsl.lsl.StreamInfo(...)
sinfo.set_channel_info(info)
info = sinfo.get_channel_info()

Please report encountered issues on the tracker.