What’s new

Using the latest features typically require being on an up-to-date mnefun main branch and often an up-to-date MNE-Python main branch.

Changelog

2021

• 2021/07/29:

  Change proj=False to proj='delayed' when epoching during covariance computation, to accord with the fact that the rejection thresholds may have been determined from projected data.

• 2021/05/03:

  Add support for returning None from p.match_fun to use default scoring, and allowing numbers to be passed to p.match_fun.

• 2021/04/08:

  Added support for allow_resample during epoching to deal with sample rate mismatches.

• 2021/04/06:

  Added support for reject_epochs_by_annot to be a regular expression.

• 2021/02/23:

  • Added acq_exclude to allow excluding folder names when searching for files on a remote acquisition machine.

  • Modified fetch_raw=True functionality to be more friendly to data being transferred to NFS mounts by not trying to set as many permissions.

• 2021/02/18:

  Expanded pick_events_cov and added pick_events_autoreject to allow control over which events are chosen for the respective epoching steps. In lieu of a function, the string “restrict” is acceptable to choose event ids contained in in_numbers. (Events for final epoching are effectively selected with “restrict”.)

2020

• 2020/12/18:

  Added erm_proj_as_esss, erm_proj_lp_cut, and erm_proj_hp_cut to support using eSSS for empty-room projectors.

• 2020/12/18:

  Added float support for params.decim.

• 2020/11/02:

  Added mnefun.repeat_coreg().

• 2020/09/28:

  Added mnefun.make_dipole_projectors().

• 2020/09/17:

  Added support for HEOG and VEOG projectors via additional projection options.

• 2020/09/16:

  Added support for the following parameters to take dict (including defaultdict) so that subject-specific parameters can be used:

  • coil_dist_limit

  • coil_gof_limit

  • coil_t_step_min

  • coil_t_window

  • rotation_limit

  • translation_limit

  • coil_bad_count_duration_limit

• 2020/09/15:

  Added support for params.filter_chpi_t_window for backward compatibility with when this was fixed to 0.2 (around Feb 2020).

• 2020/08/26:

  Optimized window length for cHPI amplitude estimation by accounting for line frequency and its harmonics. The window for a standard minimum cHPI frequency of 83 Hz should now be longer (determined by the 60 Hz line component) making head position estimation more robust.

• 2020/07/21:

  Added support for cont_hp to allow high-pass filtering (in addition to existing low-pass filtering via cont_lp; thereby allowing band-pass filtering) when computing empty-room projectors.

• 2020/07/16:

  Added support for epochs_proj='delayed' and report_params['sensor'] = dict(..., proj=False) to allow plotting sensor space responses without projection.

• 2020/07/08:

  Add support for pre_fun and post_fun hooks in reports.

• 2020/06/19

  Fixed errant picking of only data channels in preprocessed raw data files.

• 2020/05/26

  Improved mne.Report generation for comparing conditions by using sliders and grouping by sensor types.

• 2020/04/28

  Added every_other support for computing evoked data with even and odd trials.

• 2020/04/01

  Added peak-detection capability to reports for the sensor and source sections, using times='peaks'. Peaks are based on whitened gfps.

• 2020/03/06

  Added Python-based Maxwell-filter automatic bad channel detection using mf_autobad_type='python'.

• 2020/02/21

  Added Python-based head position estimation using hp_type='python'.

• 2020/02/17

  Added automated QA application acq_qa to monitor acquisition directories and generate HTML reports.

2019

• 2019/12/02

  Added get_atlas_roi_mask() and get_atlas_mapping() to aid in extracting volumetric ROIs until it’s formalized in MNE.

• 2019/12/01

  Added read_params() and YAML support for Params.

Prior

• 2015-2019/12/01

  A whole bunch of undocumented updates.

• 2015/05/01

  Added simulation code.

• 2015/05/29

  Added spherical model support.