What’s new

Version 0.11

Changelog

BUG

API

Authors

The committer list for this release is the following (preceded by number of commits):

171 Eric Larson 117 Jaakko Leppakangas

58 Jona Sassenhagen 52 Mainak Jas 46 Alexandre Gramfort 33 Denis A. Engemann 28 Teon Brooks 24 Clemens Brunner 23 Christian Brodbeck 15 Mark Wronkiewicz 10 Jean-Remi King

5 Marijn van Vliet 3 Fede Raimondo 2 Alexander Rudiuk 2 emilyps14 2 lennyvarghese 1 Marian Dovgialo

Version 0.10

Changelog

BUG

  • Fix mne.io.add_reference_channels not setting info[nchan] correctly by Federico Raimondo
  • Fix mne.stats.bonferroni_correction reject mask output to use corrected p-values by Denis Engemann
  • Fix FFT filter artifacts when using short windows in overlap-add by Eric Larson
  • Fix picking channels from forward operator could return a channel ordering different from info['chs'] by Chris Bailey
  • Fix dropping of events after downsampling stim channels by Marijn van Vliet
  • Fix scaling in :func:mne.viz.utils._setup_vmin_vmax by Jaakko Leppakangas
  • Fix order of component selection in mne.decoding.csp.CSP by Clemens Brunner

API

Authors

The committer list for this release is the following (preceded by number of commits):

273 Eric Larson 270 Jaakko Leppakangas 194 Alexandre Gramfort 128 Denis A. Engemann 114 Jona Sassenhagen 107 Mark Wronkiewicz

97 Teon Brooks 81 Lorenzo De Santis 55 Yousra Bekhti 54 Jean-Remi King 48 Romain Trachel 45 Mainak Jas 40 Alexandre Barachant 32 Marijn van Vliet 26 Jair Montoya 22 Chris Holdgraf 16 Christopher J. Bailey

7 Christian Brodbeck 5 Natalie Klein 5 Fede Raimondo 5 Alan Leggitt 5 Roan LaPlante 5 Ross Maddox 4 Dan G. Wakeman 3 Daniel McCloy 3 Daniel Strohmeier 1 Jussi Nurminen

Version 0.9

Changelog

BUG

  • Fix energy conservation for STFT with tight frames by Daniel Strohmeier
  • Fix incorrect data matrix when tfr was plotted with parameters tmin, tmax, fmin and fmax by Mainak Jas
  • Fix channel names in topomaps by Alex Gramfort
  • Fix mapping of l_trans_bandwidth (to low frequency) and h_trans_bandwidth (to high frequency) in _BaseRaw.filter by Denis Engemann
  • Fix scaling source spaces when distances have to be recomputed by Christian Brodbeck
  • Fix repeated samples in client to FieldTrip buffer by Mainak Jas and Federico Raimondo
  • Fix highpass and lowpass units read from Brainvision vhdr files by Alex Gramfort
  • Add missing attributes for BrainVision and KIT systems needed for resample by Teon Brooks
  • Fix file extensions of SSP projection files written by mne commands (from _proj.fif to -prof.fif) by Alex Gramfort
  • Generating EEG layouts no longer requires digitization points by Marijn van Vliet
  • Add missing attributes to BTI, KIT, and BrainVision by Eric Larson
  • The API change to the edf, brainvision, and egi break backwards compatibility for when importing eeg data by Teon Brooks
  • Fix bug in mne.viz.plot_topo if ylim was passed for single sensor layouts by Denis Engemann
  • Average reference projections will no longer by automatically added after applying a custom EEG reference by Marijn van Vliet
  • Fix picks argument to filter in n dimensions (affects FilterEstimator), and highpass filter in FilterEstimator by Mainak Jas
  • Fix beamformer code LCMV/DICS for CTF data with reference channels by Denis Engemann and Alex Gramfort
  • Fix scalings for bad EEG channels in mne.viz.plot_topo by Marijn van Vliet
  • Fix EGI reading when no events are present by Federico Raimondo
  • Add functionality to determine plot limits automatically or by data percentiles by Mark Wronkiewicz
  • Fix bug in mne.io.edf where the channel offsets were ommitted in the voltage calculations by Teon Brooks
  • Decouple section ordering in command-line from python interface for mne-report by Mainak Jas
  • Fix bug with ICA resetting by Denis Engemann

API

  • apply_inverse functions have a new boolean parameter prepared which saves computation time by calling prepare_inverse_operator only if it is False
  • find_events and read_events functions have a new parameter mask to set some bits to a don’t care state by Teon Brooks
  • New channels module including layouts, electrode montages, and neighbor definitions of sensors which deprecates mne.layouts by Denis Engemann
  • read_raw_brainvision, read_raw_edf, read_raw_egi all use a standard montage import by Teon Brooks
  • Fix missing calibration factors for mne.io.egi.read_raw_egi by Denis Engemann and Federico Raimondo
  • Allow multiple filename patterns as a list (e.g., *raw.fif and *-eve.fif) to be parsed by mne report in Report.parse_folder() by Mainak Jas
  • read_hsp, read_elp, and write_hsp, write_mrk were removed and made private by Teon Brooks
  • When computing the noise covariance or MNE inverse solutions, the rank is estimated empirically using more sensitive thresholds, which stabilizes results by Denis Engemann and Eric Larson and Alex Gramfort
  • Raw FIFF files can be preloaded after class instantiation using raw.preload_data()
  • Add label parameter to apply_inverse by Teon Brooks
  • Deprecated label_time_courses for in_label method in SourceEstimate by Teon Brooks
  • Deprecated as_data_frame for to_data_frame by Chris Holdgraf
  • Add transform, unit parameters to read_montage by Teon Brooks
  • Deprecated fmin, fmid, fmax in stc.plot and added clim by Mark Wronkiewicz
  • Use scipy.signal.welch instead of matplotlib.psd inside compute_raw_psd and compute_epochs_psd by Yousra Bekhti Eric Larson and Denis Engemann. As a consquence, Raw.plot_raw_psds has been deprecated.
  • Raw instances returned by mne.forward.apply_forward_raw now always have times starting from zero to be consistent with all other Raw instances. To get the former start and stop times, use raw.first_samp / raw.info['sfreq'] and raw.last_samp / raw.info['sfreq'].
  • pick_types_evoked has been deprecated in favor of evoked.pick_types.
  • Deprecated changing the sensor type of channels in rename_channels by Teon Brooks
  • CUDA is no longer initialized at module import, but only when first used.
  • add_figs_to_section and add_images_to_section now have a textbox parameter to add comments to the image by Teon Brooks
  • Deprecated iir_filter_raw for fit_iir_model_raw.
  • Add montage parameter to the create_info function to create the info using montages by Teon Brooks

Authors

The committer list for this release is the following (preceded by number of commits):

515 Eric Larson 343 Denis A. Engemann 304 Alexandre Gramfort 300 Teon Brooks 142 Mainak Jas 119 Jean-Remi King

77 Alan Leggitt 75 Marijn van Vliet 63 Chris Holdgraf 57 Yousra Bekhti 49 Mark Wronkiewicz 44 Christian Brodbeck 30 Jona Sassenhagen 29 Hari Bharadwaj 27 Clément Moutard 24 Ingoo Lee 18 Marmaduke Woodman 16 Martin Luessi 10 Jaakko Leppakangas

9 Andrew Dykstra 9 Daniel Strohmeier 7 kjs 6 Dan G. Wakeman 5 Federico Raimondo 3 Basile Pinsard 3 Christoph Dinh 3 Hafeza Anevar 2 Martin Billinger 2 Roan LaPlante 1 Manoj Kumar 1 Matt Tucker 1 Romain Trachel 1 mads jensen 1 sviter

Version 0.8

Changelog

BUG

  • Fix incorrect times attribute when stc was computed using apply_inverse after decimation at epochs stage for certain, arbitrary sample frequencies by Denis Engemann
  • Fix corner case error for step-down-in-jumps permutation test (when step-down threshold was high enough to include all clusters) by Eric Larson
  • Fix selection of total number of components via float when picking ICA sources by Denis Engemann and Qunxi Dong
  • Fix writing and reading transforms after modification in measurment info by Denis Engemann and Martin Luessi and Eric Larson
  • Fix pre-whitening / rescaling when estimating ICA on multiple channels without covariance by Denis Engemann
  • Fix ICA pre-whitening, avoid recomputation when applying ICA to new data by Denis Engemann

API

  • The minimum numpy version has been increased to 1.6 from 1.4.
  • Epochs object now has a selection attribute to track provenance of selected Epochs. The length of the drop_log attribute is now the same as the length of the original events passed to Epochs. In earlier versions it had the length of the events filtered by event_id. Epochs has also now a plot_drop_log method.
  • Deprecate Epochs.drop_picks in favor of a new method called drop_channels
  • Deprecate labels_from_parc and parc_from_labels in favor of read_labels_from_annot and write_labels_to_annot
  • The default of the new add_dist option of setup_source_space to add patch information will change from False to True in MNE-Python 0.9
  • Deprecate read_evoked and write_evoked in favor of read_evokeds and write_evokeds. read_evokeds will return all Evoked instances in a file by default.
  • Deprecate setno in favor of condition in the initialization of an Evoked instance. This affects mne.fiff.Evoked and read_evokeds, but not read_evoked.
  • Deprecate mne.fiff module, use mne.io instead e.g. mne.io.Raw instead of mne.fiff.Raw.
  • Pick functions (e.g., pick_types) are now in the mne namespace (e.g. use mne.pick_types).
  • Deprecated ICA methods specific to one container type. Use ICA.fit, ICA.get_sources ICA.apply and ICA.plot_XXX for processing Raw, Epochs and Evoked objects.
  • The default smoothing method for mne.stc_to_label will change in v0.9, and the old method is deprecated.
  • As default, for ICA the maximum number of PCA components equals the number of channels passed. The number of PCA components used to reconstruct the sensor space signals now defaults to the maximum number of PCA components estimated.

Authors

The committer list for this release is the following (preceded by number of commits):

  • 418 Denis A. Engemann
  • 284 Alexandre Gramfort
  • 242 Eric Larson
  • 155 Christian Brodbeck
  • 144 Mainak Jas
  • 49 Martin Billinger
  • 49 Andrew Dykstra
  • 44 Tal Linzen
  • 37 Dan G. Wakeman
  • 36 Martin Luessi
  • 26 Teon Brooks
  • 20 Cathy Nangini
  • 15 Hari Bharadwaj
  • 15 Roman Goj
  • 10 Ross Maddox
  • 9 Marmaduke Woodman
  • 8 Praveen Sripad
  • 8 Tanay
  • 8 Roan LaPlante
  • 5 Saket Choudhary
  • 4 Nick Ward
  • 4 Mads Jensen
  • 3 Olaf Hauk
  • 3 Brad Buran
  • 2 Daniel Strohmeier
  • 2 Federico Raimondo
  • 2 Alan Leggitt
  • 1 Jean-Remi King
  • 1 Matti Hamalainen

Version 0.7

Changelog

API

  • The pick_normal parameter for minimum norm solvers has been renamed as pick_ori and normal orientation picking is now achieved by passing the value “normal” for the pick_ori parameter.
  • ICA objects now expose the measurment info of the object fitted.
  • Average EEG reference is now added by default to Raw instances.
  • Removed deprecated read/write_stc/w, use SourceEstimate methods instead
  • The chs argument in mne.layouts.find_layout is deprecated and will be removed in MNE-Python 0.9. Use info instead.
  • plot_evoked and Epochs.plot now open a new figure by default. To plot on an existing figure please specify the axes parameter.

Authors

The committer list for this release is the following (preceded by number of commits):

  • 336 Denis A. Engemann
  • 202 Eric Larson
  • 193 Roman Goj
  • 138 Alexandre Gramfort
  • 99 Mainak Jas
  • 75 Christian Brodbeck
  • 60 Martin Luessi
  • 40 Teon Brooks
  • 29 Romain Trachel
  • 28 Andrew Dykstra
  • 12 Mark Wronkiewicz
  • 10 Christoph Dinh
  • 8 Alan Leggitt
  • 3 Yaroslav Halchenko
  • 3 Daniel Strohmeier
  • 2 Mads Jensen
  • 2 Praveen Sripad
  • 1 Luke Bloy
  • 1 Emanuele Olivetti
  • 1 Yousra BEKHTI

Version 0.6

Changelog

API

  • Deprecated use of fiff.pick_types without specifying exclude – use either [] (none), bads (bad channels), or a list of string (channel names).
  • Depth bias correction in dSPM/MNE/sLORETA make_inverse_operator is now done like in the C code using only gradiometers if present, else magnetometers, and EEG if no MEG channels are present.
  • Fixed-orientation inverse solutions need to be made using fixed=True option (using non-surface-oriented forward solutions if no depth weighting is used) to maintain compatibility with MNE C code.
  • Raw.save() will only overwrite the destination file, if it exists, if option overwrite=True is set.
  • mne.utils.set_config(), get_config(), get_config_path() moved to mne namespace.
  • Raw constructor argument proj_active deprecated – use proj argument instead.
  • Functions from the mne.mixed_norm module have been moved to the mne.inverse_sparse module.
  • Deprecate CTF compensation (keep_comp and dest_comp) in Epochs and move it to Raw with a single compensation parameter.
  • Remove artifacts module. Artifacts- and preprocessing related functions can now be found in mne.preprocessing.

Authors

The committer list for this release is the following (preceded by number of commits):

  • 340 Eric Larson
  • 330 Denis A. Engemann
  • 204 Alexandre Gramfort
  • 72 Christian Brodbeck
  • 66 Roman Goj
  • 65 Martin Luessi
  • 37 Teon Brooks
  • 18 Mainak Jas
  • 9 Simon Kornblith
  • 7 Daniel Strohmeier
  • 6 Romain Trachel
  • 5 Yousra BEKHTI
  • 5 Brad Buran
  • 1 Andrew Dykstra
  • 1 Christoph Dinh

Version 0.5

Changelog

API

  • Added nave parameter to source_induced_power() and source_band_induced_power(), use nave=1 by default (wrong nave was used before).
  • Use mne.layout.read_layout instead of mne.layout.Layout to read a layout file (.lout)
  • Use raw.time_as_index instead of time_to_index (still works but is deprecated).
  • The artifacts module (mne.artifacts) is now merged into mne.preprocessing
  • Epochs objects now also take dicts as values for the event_id argument. They now can represent multiple conditions.

Authors

The committer list for this release is the following (preceded by number of commits):

  • 313 Eric Larson
  • 226 Alexandre Gramfort
  • 219 Denis A. Engemann
  • 104 Christian Brodbeck
  • 85 Martin Luessi
  • 6 Daniel Strohmeier
  • 4 Teon Brooks
  • 1 Dan G. Wakeman

Version 0.4

Changelog

  • Add function to compute source PSD using minimum norm by Alex Gramfort
  • L21 Mixed Norm Estimates (MxNE) by Alex Gramfort and Daniel Strohmeier
  • Generation of simulated evoked responses by Alex Gramfort, Daniel Strohmeier, and Martin Luessi
  • Fit AR models to raw data for temporal whitening by Alex Gramfort.
  • speedup + reduce memory of mne.morph_data by Alex Gramfort.
  • Backporting scipy.signal.firwin2 so filtering works with old scipy by Alex Gramfort.
  • LCMV Beamformer for evoked data, single trials, and raw data by Alex Gramfort and Martin Luessi.
  • Add support for reading named channel selections by Martin Luessi.
  • Add Raw.filter method to more easily band pass data by Alex Gramfort.
  • Add tmin + tmax parameters in mne.compute_covariance to estimate noise covariance in epochs baseline without creating new epochs by Alex Gramfort.
  • Add support for sLORETA in apply_inverse, apply_inverse_raw, apply_inverse_epochs (API Change) by Alex Gramfort.
  • Add method to regularize a noise covariance by Alex Gramfort.
  • Read and write measurement info in forward and inverse operators for interactive visualization in mne_analyze by Alex Gramfort.
  • New mne_compute_proj_ecg.py and mne_compute_proj_eog.py scripts to estimate ECG/EOG PCA/SSP vectors by Alex Gramfort and Martin Luessi.
  • Wrapper function and script (mne_maxfilter.py) for Elekta Neuromag MaxFilter(TM) by Martin Luessi
  • Add method to eliminate stimulation artifacts from raw data by linear interpolation or windowing by Daniel Strohmeier.

Authors

The committer list for this release is the following (preceded by number of commits):

  • 118 Alexandre Gramfort
  • 81 Martin Luessi
  • 15 Daniel Strohmeier
  • 4 Christian Brodbeck
  • 4 Louis Thibault
  • 2 Brad Buran

Version 0.3

Changelog

  • Sign flip computation for robust label average of signed values by Alex Gramfort.
  • Reading and writing of .w files by Martin Luessi.
  • Support for modifying Raw object and allow raw data preloading with memory mapping by Martin Luessi and Alex Gramfort.
  • Support of arithmetic of Evoked data (useful to concatenate between runs and compute contrasts) by Alex Gramfort.
  • Support for computing sensor space data from a source estimate using an MNE forward solution by Martin Luessi.
  • Support of arithmetic of Covariance by Alex Gramfort.
  • Write BEM surfaces in Python by Alex Gramfort.
  • Filtering operations and apply_function interface for Raw object by Martin Luessi.
  • Support for complex valued raw fiff files and computation of analytic signal for Raw object by Martin Luessi.
  • Write inverse operators (surface and volume) by Alex Gramfort.
  • Covariance matrix computation with multiple event types by Martin Luessi.
  • New tutorial in the documentation and new classes and functions reference page by Alex Gramfort.

Authors

The committer list for this release is the following (preceded by number of commits):

  • 80 Alexandre Gramfort
  • 51 Martin Luessi

Version 0.2

Changelog

  • New stats functions for FDR correction and Bonferroni by Alex Gramfort.
  • Faster time-frequency using downsampling trick by Alex Gramfort.
  • Support for volume source spaces by Alex Gramfort (requires next MNE release or nightly).
  • Improved Epochs handling by Martin Luessi (slicing, drop_bad_epochs).
  • Bug fix in Epochs + ECG detection by Manfred Kitzbichler.
  • New pick_types_evoked function by Alex Gramfort.
  • SourceEstimate now supports algebra by Alex Gramfort.

API changes summary

Here are the code migration instructions when upgrading from mne-python version 0.1:

  • New return values for the function find_ecg_events

Authors

The committer list for this release is the following (preceded by number of commits):

  • 33 Alexandre Gramfort
  • 12 Martin Luessi
  • 2 Yaroslav Halchenko
  • 1 Manfred Kitzbichler