bd802fe | Diego Vidaurre | 03 March 2021, 22:24:58 UTC | several fixes, one particularly important, for an error introduced on 16Feb | 03 March 2021, 22:24:58 UTC |
a2089cf | Diego Vidaurre | 16 February 2021, 19:43:40 UTC | Fixed an error that made the initialisation quite inefficient (importnat one) | 16 February 2021, 19:43:40 UTC |
23528ce | Diego Vidaurre | 16 February 2021, 10:03:14 UTC | Merge branch 'master' of https://github.com/OHBA-analysis/HMM-MAR | 16 February 2021, 10:03:14 UTC |
e5f20c2 | vidaurre | 16 February 2021, 10:00:43 UTC | added checkData.m, to check the dimensions data and its conformity with T | 16 February 2021, 10:00:43 UTC |
7615c48 | Cam Higgins | 11 February 2021, 15:20:58 UTC | Added sliding window and K=1 options for baseline decoding comparisons | 11 February 2021, 15:20:58 UTC |
ff82ee0 | vidaurre | 11 February 2021, 09:46:47 UTC | tiny thing pending from last commit | 11 February 2021, 09:46:47 UTC |
1674fd1 | vidaurre | 11 February 2021, 09:38:17 UTC | Small changes in the 2016 Neuroimage example script | 11 February 2021, 09:38:17 UTC |
abc0b96 | vidaurre | 11 February 2021, 09:21:21 UTC | Fixed hmmspectramar so that it doesn't crash if T is provided in weird formats | 11 February 2021, 09:21:21 UTC |
ffa01ab | vidaurre | 11 February 2021, 09:03:15 UTC | Stop some annoying warnings | 11 February 2021, 09:03:15 UTC |
036a07c | vidaurre | 10 February 2021, 21:44:14 UTC | Fixed an error in stochastic training introduced in yesterday's commit | 10 February 2021, 21:44:14 UTC |
088d60c | vidaurre | 10 February 2021, 19:12:56 UTC | Updated the example script for the 2016 paper | 10 February 2021, 19:12:56 UTC |
89b7301 | vidaurre | 10 February 2021, 19:11:48 UTC | Improved the logic of the initialisation (now it'll converge quicker) | 10 February 2021, 19:11:48 UTC |
ff992e6 | vidaurre | 09 February 2021, 20:00:35 UTC | fixed bug introduced in the last commit | 09 February 2021, 20:00:35 UTC |
5936faa | vidaurre | 09 February 2021, 19:36:31 UTC | Lots of changes with new mode | 09 February 2021, 19:36:31 UTC |
57bc793 | vidaurre | 09 February 2021, 19:33:27 UTC | Lots of changes with new model | 09 February 2021, 19:33:27 UTC |
63d0400 | vidaurre | 09 February 2021, 19:15:37 UTC | solved issue in hmmperturb with random sampling | 09 February 2021, 19:15:37 UTC |
1694718 | vidaurre | 09 February 2021, 18:52:56 UTC | checked a recently introduced bug in checkoptions.m | 09 February 2021, 18:52:56 UTC |
12325a7 | Cam Higgins | 09 February 2021, 17:34:34 UTC | bug fix | 09 February 2021, 17:34:34 UTC |
968eee6 | Cam Higgins | 09 February 2021, 16:34:48 UTC | Merge branch 'master' of github.com:OHBA-analysis/HMM-MAR | 09 February 2021, 16:34:48 UTC |
192905d | Cam Higgins | 09 February 2021, 16:34:15 UTC | added PLS dimensionality reduction and some bug fixes | 09 February 2021, 16:34:15 UTC |
f3b5e0f | vidaurre | 05 February 2021, 15:55:08 UTC | option.plotGamma to display graphical progress during training | 05 February 2021, 15:55:08 UTC |
82fc7a9 | Diego Vidaurre | 01 February 2021, 16:53:54 UTC | Merge branch 'master' of https://github.com/OHBA-analysis/HMM-MAR | 01 February 2021, 16:53:54 UTC |
9af05cc | vidaurre | 01 February 2021, 15:57:03 UTC | fixed a typo in NatComms2018_fullpipeline.m | 01 February 2021, 15:57:03 UTC |
afffb36 | Diego Vidaurre | 01 February 2021, 07:59:03 UTC | Update README.md | 01 February 2021, 07:59:03 UTC |
1f7aad2 | vidaurre | 01 February 2021, 07:44:01 UTC | Introduced a warning about the risk of overparametrisation in the MAR model | 01 February 2021, 07:44:01 UTC |
368210e | vidaurre | 26 January 2021, 19:43:02 UTC | Scripts for my PLOS Comp Bio paper | 26 January 2021, 19:43:02 UTC |
a87bc1e | Cam Higgins | 14 January 2021, 10:09:28 UTC | Merge branch 'master' of github.com:OHBA-analysis/HMM-MAR | 14 January 2021, 10:09:28 UTC |
dd5447a | Cam Higgins | 14 January 2021, 10:09:00 UTC | Updated TUDA code to allow for supervised dimensionality reduction in cross validation procedure using PLS | 14 January 2021, 10:09:00 UTC |
f191626 | vidaurre | 12 January 2021, 18:36:59 UTC | New option BIGinitStrategy added for stochastic learning | 12 January 2021, 18:36:59 UTC |
071bce7 | vidaurre | 11 January 2021, 15:55:22 UTC | Improved some little things with the stochastic inference | 11 January 2021, 15:55:22 UTC |
adde522 | vidaurre | 03 January 2021, 12:32:51 UTC | Added a function to plot Viterbi path | 03 January 2021, 12:32:51 UTC |
7b0b4b5 | vidaurre | 03 January 2021, 12:20:34 UTC | Fixed an error when useParallel=1 and order>0 | 03 January 2021, 12:20:34 UTC |
578f6a0 | vidaurre | 22 December 2020, 20:43:06 UTC | Fixed some errors in experimental options, and improved the efficiency of hsinference a bit | 22 December 2020, 20:43:06 UTC |
e91c75e | vidaurre | 19 December 2020, 15:45:51 UTC | Fixed an issue with hmmspectramar.m | 19 December 2020, 15:45:51 UTC |
0f5813b | Cam Higgins | 14 December 2020, 08:52:07 UTC | Merge branch 'master' of github.com:OHBA-analysis/HMM-MAR | 14 December 2020, 08:52:07 UTC |
fb0ef65 | vidaurre | 04 December 2020, 18:02:34 UTC | changed the codification of the classes for TUDA | 04 December 2020, 18:02:34 UTC |
41e61c1 | Cam Higgins | 30 November 2020, 21:34:25 UTC | Minor edits to tuda configuration - specifically initialisation routine when full covariance used | 30 November 2020, 21:34:25 UTC |
b1339b1 | vidaurre | 27 November 2020, 16:29:27 UTC | Adapted padGamma to the case when the HMM is run on downsampled data | 27 November 2020, 16:29:27 UTC |
e995c30 | vidaurre | 27 November 2020, 16:21:49 UTC | Adapted padGamma to the case when the HMM is run on downsampled data (using options.downsample) | 27 November 2020, 16:21:49 UTC |
5c55fc9 | vidaurre | 24 November 2020, 15:04:33 UTC | Fixed a bug in makeMap (related to the eigenvector maps) | 24 November 2020, 15:04:33 UTC |
7dcb56b | vidaurre | 19 November 2020, 14:41:33 UTC | Removed references to squash, reordered a few files, and added predictPhenotype_CVHMM.m | 19 November 2020, 14:41:33 UTC |
6383278 | vidaurre | 13 November 2020, 16:42:13 UTC | updated some small things on the plotting functions | 13 November 2020, 16:42:13 UTC |
2d695cc | vidaurre | 09 November 2020, 09:20:49 UTC | Changed how the degree is computed in makeMap.m and fixed a small bug in hmmsinit.m | 09 November 2020, 09:20:49 UTC |
ce31fba | vidaurre | 05 November 2020, 18:11:33 UTC | Improved the prediction stuff | 05 November 2020, 18:11:33 UTC |
5fe3a62 | vidaurre | 04 November 2020, 14:29:16 UTC | corrected bug in hmm_kl.m | 04 November 2020, 14:29:16 UTC |
56ecdaf | vidaurre | 04 November 2020, 14:28:50 UTC | changed value by default of BIGverbose | 04 November 2020, 14:28:50 UTC |
eb339cb | vidaurre | 04 November 2020, 14:28:18 UTC | corrected bug in getGammaSimilarity.m | 04 November 2020, 14:28:18 UTC |
78c60f4 | vidaurre | 03 November 2020, 20:03:36 UTC | Make the stochastic inference to run with mixture models | 03 November 2020, 20:03:36 UTC |
7d5d7ed | vidaurre | 03 November 2020, 20:03:14 UTC | Updated the prediction stuff | 03 November 2020, 20:03:14 UTC |
5879dea | vidaurre | 02 November 2020, 10:57:19 UTC | continuation of last commitm | 02 November 2020, 10:57:19 UTC |
ee9edee | vidaurre | 02 November 2020, 10:28:17 UTC | Expanded the functionality of the plotting functions and add some utilities to handle parcellations | 02 November 2020, 10:28:17 UTC |
56a8b2c | Cam Higgins | 27 October 2020, 13:53:52 UTC | Added option to return both COD and Pearson accuracy metrics from call to tudacv | 27 October 2020, 13:53:52 UTC |
132bfb7 | Cam Higgins | 27 October 2020, 13:53:09 UTC | Minor improvements to LDA prediction code | 27 October 2020, 13:53:09 UTC |
e43057c | vidaurre | 27 October 2020, 12:36:52 UTC | fixed something that could give trouble when the HMM on a clusterM | 27 October 2020, 12:36:52 UTC |
2172687 | vidaurre | 21 October 2020, 15:38:53 UTC | Large update including a new model based on the factorial HMM | 21 October 2020, 15:38:53 UTC |
27ded33 | vidaurre | 21 October 2020, 14:07:32 UTC | more minor updates for stochastic learning | 21 October 2020, 14:07:32 UTC |
1283d3e | vidaurre | 21 October 2020, 14:05:26 UTC | minor changes related to the stochastic learning option | 21 October 2020, 14:05:26 UTC |
568cf54 | vidaurre | 21 October 2020, 14:02:41 UTC | added function to pass parcellations from func.gii to nifti | 21 October 2020, 14:02:41 UTC |
64e6bc5 | Administrator | 19 October 2020, 12:21:41 UTC | Merge branch 'master' of https://github.com/OHBA-analysis/HMM-MAR | 19 October 2020, 12:21:41 UTC |
a6fd548 | vidaurre | 19 October 2020, 12:19:11 UTC | Solved small issues with brain maps | 19 October 2020, 12:19:11 UTC |
784adde | Cam Higgins | 16 October 2020, 17:10:15 UTC | Merge branch 'master' of github.com:OHBA-analysis/HMM-MAR | 16 October 2020, 17:10:15 UTC |
c336cb0 | Cam Higgins | 16 October 2020, 17:09:57 UTC | Optimisation of LDA inference code | 16 October 2020, 17:09:57 UTC |
b36bf13 | vidaurre | 16 October 2020, 15:30:40 UTC | Make possible to produce brain maps for some states only | 16 October 2020, 15:30:40 UTC |
c37f3dd | vidaurre | 16 October 2020, 09:31:01 UTC | adapt NatComms2018_fullpipeline.m to use the new connectivity plotting functionsm | 16 October 2020, 09:31:01 UTC |
3cfafd9 | vidaurre | 16 October 2020, 09:27:14 UTC | Rebalance the parcellations in the brain map projections | 16 October 2020, 09:27:14 UTC |
39dc2ce | vidaurre | 16 October 2020, 09:16:37 UTC | Small edits in the brain map functions | 16 October 2020, 09:16:37 UTC |
c81b1d2 | vidaurre | 16 October 2020, 08:37:10 UTC | Added functionality to create brain activity and connectivity from spectral estimations | 16 October 2020, 08:37:10 UTC |
bd973e4 | Administrator | 15 October 2020, 14:54:26 UTC | Added functions to display the HMM states into brain activity and connectivity maps | 15 October 2020, 14:54:26 UTC |
dd224a3 | Administrator | 15 October 2020, 14:53:33 UTC | Improved the stochastic inference | 15 October 2020, 14:53:33 UTC |
9fdea97 | vidaurre | 08 October 2020, 19:42:15 UTC | Huge improve in memory usage when options.useParallel is used (3) | 08 October 2020, 19:42:15 UTC |
48f76b7 | vidaurre | 08 October 2020, 17:15:45 UTC | Huge improve in memory usage when options.useParallel is used (2) | 08 October 2020, 17:15:45 UTC |
f31d04f | Administrator | 08 October 2020, 16:06:33 UTC | Fixed issue with stochastic learning when covtype='uniquediag'/'uniquefull' | 08 October 2020, 16:06:33 UTC |
166966b | Administrator | 08 October 2020, 16:05:10 UTC | Huge improve in memory usage when options.useParallel is used | 08 October 2020, 16:05:10 UTC |
4b946d1 | Administrator | 05 October 2020, 10:41:33 UTC | Corrected small typo in getFuncConn.m | 05 October 2020, 10:41:33 UTC |
f079f84 | vidaurre | 28 September 2020, 10:03:30 UTC | Solved rank-related issue with getFuncConn.m when PCA is used | 28 September 2020, 10:03:30 UTC |
1372b1c | vidaurre | 25 September 2020, 11:59:45 UTC | fixed a bug in getFuncConn.m | 25 September 2020, 11:59:45 UTC |
c56b92b | vidaurre | 25 September 2020, 08:59:47 UTC | getFuncConn.m now returns also partial correlation matrices | 25 September 2020, 08:59:47 UTC |
63e8094 | vidaurre | 18 September 2020, 13:15:27 UTC | Option on getFuncConn.m and getMean.m to return parameters in PCA space | 18 September 2020, 13:15:27 UTC |
4534d1c | Cam Higgins | 11 September 2020, 10:54:09 UTC | Fixing tuda/tuca outputs to use soft probabilities | 11 September 2020, 10:54:09 UTC |
8a28277 | Cam Higgins | 11 September 2020, 10:40:05 UTC | Fixed precision bug in state likelihood computation | 11 September 2020, 10:40:05 UTC |
a066945 | Cam Higgins | 11 September 2020, 08:59:25 UTC | Merge remote-tracking branch 'refs/remotes/github/master' | 11 September 2020, 08:59:25 UTC |
3551863 | Cam Higgins | 11 September 2020, 08:58:40 UTC | Added AUC and cross-validated model likelihood as classification accuracy metrics | 11 September 2020, 08:58:40 UTC |
8f97ef3 | Cam Higgins | 11 September 2020, 08:30:52 UTC | Amended to return soft classification probabilities | 11 September 2020, 08:30:52 UTC |
b85d1b4 | Cam Higgins | 11 September 2020, 08:23:44 UTC | Improved intialisation for encoding models | 11 September 2020, 08:23:44 UTC |
8cf047b | Cam Higgins | 11 September 2020, 07:48:17 UTC | Updated tudadecode to work with encoding models | 11 September 2020, 07:48:17 UTC |
7c6e9e0 | vidaurre | 03 September 2020, 15:24:42 UTC | Merge branch 'master' of https://github.com/OHBA-analysis/HMM-MAR | 03 September 2020, 15:24:42 UTC |
be88157 | vidaurre | 03 September 2020, 15:17:39 UTC | add functionality for HMM-PCA | 03 September 2020, 15:17:39 UTC |
0486e27 | vidaurre | 03 September 2020, 15:17:21 UTC | simplify cvhmmmar.m | 03 September 2020, 15:17:21 UTC |
1091377 | Cam Higgins | 25 August 2020, 13:56:26 UTC | Merge remote-tracking branch 'refs/remotes/github/master' | 25 August 2020, 13:56:26 UTC |
189f836 | Cam Higgins | 25 August 2020, 13:55:47 UTC | Fix bug and enforce hard state assignments when computing state conditioned accuracies in tuda | 25 August 2020, 13:55:47 UTC |
8ddb798 | Cam Higgins | 25 August 2020, 13:10:47 UTC | Update coefficient of determination computation for tudacv | 25 August 2020, 13:10:47 UTC |
673ed58 | vidaurre | 24 August 2020, 15:58:03 UTC | Merge branch 'master' of https://github.com/OHBA-analysis/HMM-MAR | 24 August 2020, 15:58:03 UTC |
2b432b1 | vidaurre | 24 August 2020, 15:57:39 UTC | Remove PCA message when verbose=0 | 24 August 2020, 15:57:39 UTC |
5ba8a80 | Cam Higgins | 24 August 2020, 14:33:25 UTC | Merge branch 'master' of github.com:OHBA-analysis/HMM-MAR | 24 August 2020, 14:33:25 UTC |
7506f82 | Cam Higgins | 24 August 2020, 14:32:36 UTC | Couple of TUDA bug fixes for cases where more than one regressor is used | 24 August 2020, 14:32:36 UTC |
b52073b | vidaurre | 24 August 2020, 14:06:11 UTC | Optimized HMM-PCA for computational efficiency | 24 August 2020, 14:06:11 UTC |
3cb7315 | vidaurre | 24 August 2020, 13:47:14 UTC | Optimized calls to setstateoptions.m to improve computational efficiency | 24 August 2020, 13:47:14 UTC |
dd0a8e4 | Cam Higgins | 14 August 2020, 17:11:05 UTC | Added option in TUDA to return accuracy per state (averaged over all state activations | 14 August 2020, 17:11:05 UTC |
5391b94 | Cam Higgins | 14 August 2020, 17:09:54 UTC | Fixed bug when doing binary classification with LDA | 14 August 2020, 17:09:54 UTC |
0ccc302 | Cam Higgins | 13 August 2020, 16:28:54 UTC | For decorrelating regressors, ensured outputs are scaled to unit variance | 13 August 2020, 16:28:54 UTC |