What’s new?#
Version 0.11.1 (2022-10-21)#
Version 0.11.1 is a patch release, all changes are listed below. For more complete information on the 0.11 version, see changelog below.
Speed up
mne_bids.read_raw_bids()when lots of events are present by Alexandre Gramfort (#1079)When writing data containing
mne.Annotationsand passing events towrite_raw_bids(), previously, annotations whose description did not appear inevent_idwere silently dropped. We now raise an exception and request users to specify mappings between descriptions and event codes in this case. It is still possible to omitevent_idif noeventsare passed, by Richard Höchenberger (#1084)
Version 0.11 (2022-10-08)#
This release includes a number of bug fixes as well as several smaller enhancements. Please note some behavior changes and updated requirements, as listed in the details below.
📝 Notable changes#
Support for new channel types is available: temperature and galvanic skin response
MNE-BIDS now supports the BIDS “description” entity (
desc)It’s now possible to store Curry (CDT) files and EGI files with MNE-BIDS.
Detailed list of changes#
🚀 Enhancements#
You can now write raw data and an associated empty-room recording with just a single call to
mne_bids.write_raw_bids(): theempty_roomparameter now also accepts anmne.io.Rawdata object. The empty-room session name will be derived from the recording date automatically, by Richard Höchenberger (#998)write_raw_bids()now stores participant weight and height inparticipants.tsv, by Richard Höchenberger (#1031)write_raw_bids()now supports EGI format, by Anand Saini, Scott Huberty and Mathieu Scheltienne (#1006)When a given subject cannot be found, valid suggestions are now printed, by Eric Larson (#1066)
TSV files that are empty (i.e., only a header row is present) are now handled more robustly and a warning is issued, by Stefan Appelhoff (#1038)
BIDSPathnow supports the BIDS “description” entitydesc, used in derivative data, by Richard Höchenberger (#1049)Added support for
GSR(galvanic skin response / electrodermal activity, EDA) andTEMP(temperature) channel types, by Richard Höchenberger (#1059)Added support for reading EEG files in Curry 8 format (‘.cdt’ extension) by Denis Engemann (#1072)
🧐 API and behavior changes#
write_raw_bids()now expects all but the first four parameters to be passed as keyword arguments, by Richard Höchenberger (#1054)The
events_dataparameter ofwrite_raw_bids()has been deprecated in favor of a new parameter namedevents. This ensures more consistency between the MNE-BIDS and MNE-Python APIs. You may continue using theevents_dataparameter for now, but aFutureWarningwill be raised.events_datawill be removed in MNE-BIDS 0.14, by Richard Höchenberger (#1054)In many places, we used to infer the
datatypeof aBIDSPathfrom thesuffix, if not explicitly provided. However, this has lead to trouble in certain edge cases. In an effort to reduce the amount of implicit behavior in MNE-BIDS, we now require users to explicitly specify adatatypewhenever the invoked functions or methods expect one, by Richard Höchenberger (#1030)mne_bids.make_dataset_description()now accepts keyword arguments only, and can now also write the following metadata:HEDVersion,EthicsApprovals,GeneratedBy, andSourceDatasets, by Stefan Appelhoff (#406)The deprecated function
mne_bids.mark_bad_channelshas been removed in favor ofmne_bids.mark_channels(), by Richard Höchenberger (#1009)mne_bids.print_dir_tree()now raises aFileNotFoundErrorinstead of aValueErrorif the directory does not exist, by Richard Höchenberger (#1013)Passing only one of
eventsandevent_idtowrite_raw_bids()now raises aValueErrorinstead of aRuntimeError, by Richard Höchenberger (#1054)Until now,
mne_bids.BIDSPathprepends extensions with a period “.” automatically. We intend to remove this undocumented side-effect and now emit aFutureWarningif anextensionthat does not start with a.is provided. Starting with MNE-BIDS 0.12, an exception will be raised in this case, by Richard Höchenberger (#1061)Provide a more helpful error message when trying to write non-preloaded concatenated data, by Richard Höchenberger (##1075)
🛠 Requirements#
MNE-BIDS now requires MNE-Python 1.0 or newer.
Writing BrainVision files now requires
pybvversion0.7.3, by Stefan Appelhoff (#1011)
🪲 Bug fixes#
Fix ACPC in
surface RASinstead ofscanner RASin 08. Convert iEEG data to BIDS format and add convenience functionsmne_bids.convert_montage_to_ras()andmne_bids.convert_montage_to_mri()to help, by Alex Rockhill (#990)Suppress superfluous warnings about MaxShield in many functions when handling Elekta/Neuromag/MEGIN data, by Richard Höchenberger (#1000)
The MNE-BIDS Inspector didn’t work if
mne-qt-browserwas installed and used as the default plotting backend, as the Inspector currently only supports the Matplotlib backend, by Richard Höchenberger (#1007)copyfile_brainvision()can now deal with.datfile extension, by Dominik Welke (#1008)print_dir_tree()now correctly expands~to the user’s home directory, by Richard Höchenberger (#1013)write_raw_bids()now correctly excludes stim channels when writing to electrodes.tsv, by Scott Huberty (#1023)read_raw_bids()doesn’t populateraw.info['subject_info']with invalid values anymore, preventing users from writing the data to disk again, by Richard Höchenberger (#1031)Writing EEGLAB files was sometimes broken when
.setand.fdtpairs were supplied. This is now fixed incopyfile_eeglab(), by Stefan Appelhoff (#1039)Writing and copying CTF files now works on Windows when files already exist (
overwrite=True), by Stefan Appelhoff (#1035)Instead of deleting files and raising cryptic errors, an intentional error message is now sent when calling
write_raw_bids()with the source file identical to the destination file, unlessformatis specified, by Adam Li and Stefan Appelhoff (#889)Internal helper function to
read_raw_bids()would reject BrainVision data if_scans.tsvlisted a.eegfile instead of.vhdr, by Teon Brooks (#1034)Whenever
read_raw_bids()encounters a channel type that currently doesn’t translate into an appropriate MNE channel type, the channel type will now be set to'misc. Previously, seemingly arbitrary channel types would be applied, e.g.'eeg'for GSR and temperature channels, by Richard Höchenberger (#1052)Fix the incorrect setting of the fields
ContinuousHeadLocalizationandHeadCoilFrequencyfor Neuromag MEG recordings, by Eduard Ort (#1067)