https://github.com/MICA-MNI/BrainSpace
Tip revision: 91409586d68c290108d0eee6de78cccee5a8f6dc authored by Reinder Vos de Wael on 24 August 2020, 22:15:00 UTC
[ENH] Added parallel pool support.
[ENH] Added parallel pool support.
Tip revision: 9140958
CHANGELOG.txt
v0.1.2, XXXXX 2020 -- Let there be null models update
MATLAB
- New functionality
- Added a verbose flag to GradientMaps (default value: false).
- plot_hemispheres has several new functionalities:
- Now returns an object rather than the graphics handles.
Old graphics handles are stored in object.handles.
- Added the following methods to the plot_hemispheres object for
modifying figure properties:
- object.colorlimits: modifies color limits.
- object.colormaps: modifies colormaps.
- object.labels: modifies labeltext properties.
- Added a "views" name-value pair that enables anterior,
posterior, superior, and inferior views.
ReadTheDocs
- Updated MATLAB documentation of GradientMaps with the verbose flag.
- Updated MATLAB documentation of plot_hemispheres to reflect the new
changes.
v0.1.1, March 2020 -- Quality of life update
MATLAB:
- New functionality:
- split_surfaces.m
- surface_to_graph.m
- load_group_mpc.m
- Bug fixes:
- Fixed a bug where the example data loaders failed when the BrainSpace
directory was not named "BrainSpace".
- Matrix sparsification now correctly thresholds -inf.
- The "Running with sparsity parameter..." message will now correctly
display with every kernel.
- Fixed a bug where joint embedding of non-square matrices would fail.
- plot_hemispheres will no longer error when using non-float data.
- Other changes:
- Optimized the spin_permutation code. It should now run substantially
faster, especially with a large number of permutations.
- Improved several warning/error messages.
- Moved execution of custom kernels to after the sparsification.
Python:
- gradient:
- Fix bug in alignment (remove centering and scaling)
- Add support for procrutes alignment of one subject to reference in
GradientMaps
- Remove dependency on _graph_is_connected from scikit-learn
- datasets:
- Add load_group_mpc
- mesh:
- Update build_polydata with new vtk_interface
- Add downsample_with_parcellation to mesh_operations
- Fixed a bug with mask for smooth_array and add support for
multicomponent arrays (columns). Smooth columns separately.
- parcellation
- Fixed a bug in find_label_correspondence
- plotting:
- Add plotting defaults
- Add build_plotter
- Update plot_surf
- Remove numpy warning plot_hemispheres
- vtk_interface:
- Reorganize and add several wrappers
- Update wrap_vtk to look for superclasses is BS wrapper doesn't exist
- Add support for vtkSringArray and vtkVariantArray
- Add vtype kwarg in decorators for unwarpping numpy to vtk abstractarray
- Add support for dict arguments
- Add automatic wrap/unwrap output/input of vtk methods
- Add new repr
- Add wrappers BSActor2D, BSScalarBarActor, BSTexturedActor2D, BSTextActor
- Add AddScalarBarActor and AddTextActor methods for BSViewport
- Add io_support from mesh here
- tests:
- Update tests with new vtk_interface
- requiremnts:
- Update to scikit-learn>=0.22.0
- Remove pillow
ReadTheDocs:
- Added matrix fusion example to MATLAB tutorial 2.
v0.1.0, September 2019 -- Initial release