Version 1.4.0 (2023-05-10)

Enhancements

• Add functionality for reading CNT spans/annotations marked bad to mne.io.read_raw_cnt() (#11631 by new contributor Jacob Woessner)

• Add :unit: Sphinx directive to enable use of uniform non-breaking spaces throughout the documentation (#11469 by new contributor Sawradip Saha)

• Adjusted the algorithm used in mne.decoding.SSD to support non-full rank data (#11458 by new contributor Thomas Binns)

• Changed suggested type for ch_groups in mne.viz.plot_sensors from array to list of list(s) (arrays are still supported). (#11465 by Hyonyoung Shin)

• Add support for UCL/FIL OPM data using mne.io.read_raw_fil() (#11366 by new contributor George O’Neill and Robert Seymour)

• Add harmonic field correction (HFC) for OPM sensors in mne.preprocessing.compute_proj_hfc() (#11536 by new contributor George O’Neill and Eric Larson)

• Forward argument axes from mne.viz.plot_sensors to mne.channels.DigMontage.plot (#11470 by new contributor Jan Ebert and Mathieu Scheltienne)

• Add forward IIR filtering, using parameters method='iir', phase='forward' (#11078 by new contributor Quentin Barthélemy)

• Added ability to read stimulus durations from SNIRF files when using mne.io.read_raw_snirf() (#11397 by Robert Luke)

• Add mne.Info.save() to save an mne.Info object to a fif file (#11401 by Alex Rockhill)

• Improved error message when downloads are corrupted for mne.datasets.sample.data_path() and related functions (#11407 by Eric Larson)

• Improved mne.get_config("") behavior to return a dict of key-value pairs (config key and its description) rather than just a tuple of config keys (#11166 by Clemens Brunner, Daniel McCloy, and Eric Larson)

• Add support for skip_by_annotation in mne.io.Raw.notch_filter() (#11388 by Mainak Jas)

• Add support for output='complex' to mne.time_frequency.psd_array_welch() and when using method='welch' with mne.Epochs.compute_psd() (#11556 by Eric Larson)

• Slightly adjusted the window title for mne.Epochs.plot() (#11419 by Richard Höchenberger and Daniel McCloy)

• Add mne.count_events() to count unique event types in a given event array (#11430 by Clemens Brunner)

• Add a video to How MNE uses FreeSurfer’s outputs of a brain inflating from the pial surface to aid in understanding the inflated brain (#11440 by Alex Rockhill)

• Add automatic projection of sEEG contact onto the inflated surface for mne.viz.Brain.add_sensors() (#11436 by Alex Rockhill)

• Allow an image with intracranial electrode contacts (e.g. computed tomography) to be used without the freesurfer recon-all surfaces to locate contacts so that it doesn’t have to be downsampled to freesurfer dimensions (for microelectrodes) and show an example Locating micro-scale intracranial electrode contacts with mne.transforms.apply_volume_registration_points() added to aid this transform (#11567 by Alex Rockhill)

• Use new dipy.align.imwarp.DiffeomorphicMap.transform_points() to transform a montage of intracranial contacts more efficiently (#11572 by Alex Rockhill)

• Improve performance of raw data browsing with many annotations (#11614 by Eric Larson)

• Add support for mne.preprocessing.maxwell_filter() with gradient-compensated CTF data, e.g., for tSSS-only mode (#10554 by Eric Larson)

• Add support for eyetracking data using mne.io.read_raw_eyelink() (#11152 by Dominik Welke and Scott Huberty)

• mne.channels.make_1020_channel_selections() gained a new parameter, return_ch_names, to allow for easy retrieval of EEG channel names corresponding to the left, right, and midline portions of the montage (#11632 by Richard Höchenberger)

• Methods for setting the sensor types of channels (e.g., for raw data, mne.io.Raw.set_channel_types()) gained a new parameter, on_unit_change, to control behavior (raise an exception, emit a warning, or do nothing) in case the measurement unit is adjusted automatically (#11668 by Richard Höchenberger)

• mne.beamformer.trap_music() implements the TRAP-MUSIC localisation algorithm with the same signature as mne.beamformer.rap_music() (#11679 by Théodore Papadopoulo)

Bugs

• Improving compatibility with Qt6 by removing the use of deprecated AA_UseHighDpiPixmaps attribute for this Qt version (#11662 by new contributor Florin Pop)

• Fix mne.time_frequency.psd_array_multitaper() docstring where argument bandwidth incorrectly reported argument as half-bandwidth and gave wrong explanation of default value (#11479 by new contributor Tom Stone)

• Fix bug where installation of a package depending on mne will error when done in an environment where setuptools is not present (#11454 by new contributor Arne Pelzer)

• Fix bug in mne.Annotations.rename() where replacements were not done correctly (#11666 by new contributor Timur Sokhin and Eric Larson)

• Fix mne.time_frequency.Spectrum.to_data_frame()’s docstring to reflect the correct name for the appended frequencies column (#11457 by new contributor Zvi Baratz)

• Fix bug with mne.io.read_raw_egi() where events outside the recording window could break data reading (#11505 by new contributor Niels Focke)

• Fix bug where mne.preprocessing.regress_artifact() and mne.preprocessing.EOGRegression incorrectly tracked picks (#11366 by Eric Larson)

• Fix bug where channel names were not properly sanitized in mne.write_evokeds() and related functions (#11399 by Eric Larson)

• Fix bug where splash screen would not always disappear (#11398 by Eric Larson)

• Fix bug where having a different combination of volumes loaded into freeview caused different affines to be returned by mne.read_lta() for the same Linear Transform Array (LTA) (#11402 by Alex Rockhill)

• Fix how mne.channels.DigMontage is set when using mne.gui.locate_ieeg so that mne.Info.get_montage() works and does not return None (#11421 by Alex Rockhill)

• Fix mne.io.read_raw_edf() when reading EDF data with different sampling rates and a mix of data channels when using infer_types=True (#11427 by Alex Gramfort)

• Fix how mne.channels.DigMontage is set when using mne.preprocessing.ieeg.project_sensors_onto_brain() so that mne.Info.get_montage() works and does not return None (#11436 by Alex Rockhill)

• Fix configuration folder discovery on Windows, which would fail in certain edge cases; and produce a helpful error message if discovery still fails (#11441 by Richard Höchenberger)

• Make SlidingEstimator and GeneralizingEstimator respect the verbose argument. Now with verbose=False, the progress bar is not shown during fitting, scoring, etc. (#11450 by Mikołaj Magnuski)

• Fix bug with mne.gui.locate_ieeg where Freesurfer ?h.pial.T1 was not recognized and suppress excess logging (#11489 by Alex Rockhill)

• All functions accepting paths can now correctly handle Path as input. Historically, we expected strings (instead of “proper” path objects), and only added Path support in a few select places, leading to inconsistent behavior. (#11473 and #11499 by Mathieu Scheltienne)

• Fix visualization dialog compatibility with matplotlib 3.7 (#11409 by Daniel McCloy and Eric Larson)

• Expand tilde (user directory) in config keys (#11537 by Clemens Brunner)

• Fix bug in mne.preprocessing.compute_maxwell_basis() where using int_order=0 would raise an error (#11562 by Eric Larson)

• In the legacy function mne.viz.plot_epochs_psd_topomap(), the parameter names now works again (#11563 by Daniel McCloy)

• Fix mne.io.read_raw() for file names containing multiple dots (#11521 by Clemens Brunner)

• Fix bug in mne.export.export_raw() when exporting to EDF with a physical range set smaller than the data range (#11569 by Mathieu Scheltienne)

• Fix bug in mne.concatenate_raws() where two raws could not be merged if the order of the bad channel lists did not match (#11502 by Moritz Gerster)

• Fix bug where mne.Evoked.plot_topomap() opened an extra figure (#11607 by Alex Rockhill)

• Fix bug where mne.transforms.apply_volume_registration_points() modified info in place (#11612 by Alex Rockhill)

• Fix bug where Maxwell-filtered data rank was not handled properly in mne.beamformer.make_lcmv() (#11664 by Eric Larson)

• In Report, custom figures now show up correctly when image_format='svg' is requested (#11623 by Richard Höchenberger)

• Fix bug where providing axes in mne.preprocessing.ICA.plot_components would fail (#11654 by Mathieu Scheltienne)

• Fix deprecation of title, vmin and vmax in plot_tfr_topomap called from interactive mne.time_frequency.AverageTFR.plot_topo (#11683 by Mathieu Scheltienne)

API changes

• Deprecate mne.gui.locate_ieeg in favor of mne_gui_addons.locate_ieeg() (#11616 by Alex Rockhill)

• Deprecate arguments kind and path from mne.channels.read_layout() in favor of a common argument fname (#11500 by Mathieu Scheltienne)

• Change aligned_ct positional argument in mne.gui.locate_ieeg to base_image to reflect that this can now be used with unaligned images (#11567 by Alex Rockhill)

• mne.warp_montage_volume was deprecated in favor of mne.preprocessing.ieeg.warp_montage() (acts directly on points instead of using an intermediate volume) and mne.preprocessing.ieeg.make_montage_volume() (which makes a volume of ieeg contact locations which can still be useful) (#11572 by Alex Rockhill)

• Deprecate mne.pick_channels_evoked in favor of evoked.copy().pick(...) (#11665 by Eric Larson)

• Set instance methods inst.pick_types and inst.pick_channels as legacy in favor of inst.pick(...) (#11665 by Eric Larson)

• The default of inst.pick_channels(..., ordered=False) will change to ordered=True in 1.5 to avoid silent bugs (#11665 by Eric Larson)

Authors

• Alex Rockhill

• Alexandre Gramfort

• Arne Pelzer+

• Britta Westner

• Carlos de la Torre

• Clemens Brunner

• Daniel Hasegan

• Daniel McCloy

• Denis A. Engemann

• Dimitri Papadopoulos Orfanos

• Dinara Issagaliyeva

• Dominik Welke

• Enzo Altamiranda

• Eric Brayet

• Eric Larson

• Erkka Heinila

• Felix Klotzsche

• Florin Pop+

• George O’Neill+

• Giulio

• Hakimeh Aslsardroud

• Hyonyoung Shin+

• Jacob Woessner+

• Jan Ebert+

• Jennifer Behnke

• Jevri Hanna

• Lukas Hecker

• Mainak Jas

• Mark Alexander Henney

• Mathieu Scheltienne

• Mauricio Cespedes

• Michiru Kaneda

• Mikołaj Magnuski

• Moritz Gerster

• Niels Focke+

• Omer Shubi

• Pavel Navratil

• Quentin Barthélemy+

• Richard Höchenberger

• Rob Luke

• Robert Seymour+

• Santeri Ruuskanen

• Sawradip Saha+

• Scott Huberty

• Stefan Appelhoff

• Theodore Papadopoulo

• Thomas Samuel Binns+

• Timon Merk

• Timur Sokhin+

• Tom Ma

• Tom Stone+

• Toomas Erik Anijärv

• Zvi Baratz+