forked from CoSMoMVPA/CoSMoMVPA
-
Notifications
You must be signed in to change notification settings - Fork 0
/
Changelog
65 lines (39 loc) · 2.94 KB
/
Changelog
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
===================
CoSMoMVPA changelog
===================
* 1.1.0 (2016-12-06)
* Fixed
* **Major** Fix issue in :ref:`cosmo_montecarlo_cluster_stat`, which returned z-scores corresponding than one-tailed instead of two-tailed probabilities, doubling the Type I error rate.
* When using GNU Octave, let :ref:`cosmo_check_external` deal properly with packages that are installed but not loaded.
* Changed behaviour
* In :ref:`cosmo_crossvalidate`, the first output contains (for each sample) a prediction for each fold, rather than one single prediction computed by :ref:`cosmo_winner_indices`. (Unchanged is that :ref:`cosmo_crossvalidation_measure` returns, by default, a single prediction per sample).
* Deprecated
* In :ref:`cosmo_crossvalidation_measure`, the "predictions" and "raw" output options are deprecated; use "winner_predictions" instead.
* :ref:`cosmo_dataset_slice_fa` and :ref:`cosmo_dataset_slice_fa` will be removed in the future; use :ref:`cosmo_slice` instead.
* :ref:`cosmo_dim_slice` will be removed in the future; use :ref:`cosmo_slice` and :ref:`cosmo_dim_prune` instead.
* :ref:`cosmo_meta_feature_selection_classifier` will be removed in the future; use :ref:`cosmo_classify_meta_feature_selection` instead.
* Removed
* in :ref:`cosmo_crossvalidation_measure`, the "accuracy_by_chunk" output option has been removed. Use "fold_predictions" to get predictions for each fold.
* in :ref:`cosmo_naive_bayes_classifier_searchlight`, the "predictions" output option has been removed. Use "winner_predictions" instead.
* New features
* :ref:`cosmo_meeg_dataset` and :ref:`cosmo_map2meeg` provide input/output support for EEGLAB datasets with study structure, and for FieldTrip datasets with "subj" as the first dimension in dimord.
* :ref:`cosmo_montecarlo_cluster_stat` and :ref:`cosmo_searchlight` support parallellization over multiple CPU cores/threads using the "nproc" parameter.
* New function :ref:`cosmo_independent_samples_partitioner` for MEEG datasets where samples can be assumed to be independent.
* New function :ref:`cosmo_randperm` with optional seed.
* New functions for parallel computations on the Matlab and Octave platform, :ref:`cosmo_parallel_get_nproc_available` and :ref:`cosmo_parcellfun`.
* New functions :ref:`cosmo_map_pca` and :ref:`cosmo_pca` for Principal Component Analysis (PCA), and support for PCA in :ref:`cosmo_crossvalidate` and :ref:`cosmo_crossvalidation_measure`.
* Added Changelog
* Contributors
* Tijl Grootswagers
* Acknowledgements
* Christoph Braun
* Tijl Grootswagers
* Thomas Hinault
* Maria Montchal
* Daria Proklova
* Joscha Schmiedt
* Jo Taylor
* Lara Todorova
* Raffaele Tucciarelli
* 1.0.0 (2016-07-22)
* First public release.