What’s new#

v0.4.6#

Bug fixes#

Fix frites.conn.conn_spec() in multitaper mode : the cross- and auto-spectra are now averaged over tapers (the complex coefficients used to be averaged before, cancelling most of the signal). Coherence and PLV values change for every multitaper user (brainets/frites#69)
Fix the hanning smoothing kernel of frites.conn.conn_spec() (sm_times of 2 samples gave a NaN output, 3 samples no smoothing) (brainets/frites#69)
The cross-spectrum (metric='sxy') of frites.conn.conn_spec() is now complex (the imaginary part used to be silently dropped) (brainets/frites#69)
frites.conn.conn_spec() sets zero_mean=False explicitly in the time-frequency decomposition : mne changed its default to True, which silently changed the results between mne versions (brainets/frites#70)
frites.set_mpl_style() no longer depends on pkg_resources (absent from Python >= 3.12 environments) (brainets/frites#70)
Remove the np.in1d compatibility shim added in v0.4.5 : no supported mne release uses np.in1d and the shim modified the NumPy namespace globally
frites.stats.trial_swap_surrogates() ignored its random_state (the trials were shuffled with the global random generator) ; surrogates are now reproducible
New random_state parameter for frites.simulations.sim_local_ccd_ms() (the conditions were drawn from the global random generator)
The copula normalization (frites.core.copnorm_nd() and the GCMI estimators built on it) now raises a ValueError on NaN values instead of silently ranking them and returning a finite, meaningless mutual information
frites.workflow.WfMi and frites.workflow.WfConnComod raise an explicit ValueError when n_perm=0 is combined with a permutation-based correction (used to fail with a bare AssertionError) ; use mcp='noperm' to only compute the mutual information
frites.conn.conn_reshape_undirected() and frites.conn.conn_reshape_directed() no longer rely on deprecated xarray behaviours (xr.concat default coords and implicit pandas.MultiIndex promotion) that were about to change their results
frites.conn.conn_spec() no longer parallelizes both the time-frequency decomposition and the loop over pairs (nested parallelism oversubscribed the CPU with the default n_jobs=-1)
frites.plot.plot_conn_circle() uses Colormap.with_extremes (set_bad is pending deprecation in matplotlib)

Dependencies#

Python >= 3.10 is now required (python_requires) ; the continuous integration matrix moved from Python 3.8/3.9 to 3.10/3.11/3.12
Tested against mne 1.5 up to 1.13 (mne >= 1.13 requires Python >= 3.11) and NumPy 2.x up to 2.5
Documentation build : sphinxcontrib-bibtex >= 2.6 and sphinx >= 7.4 (brainets/frites#70)
xarray >= 2023.8 (xarray.Coordinates.from_pandas_multiindex)

v0.4.5#

Bug fixes#

Compatibility with NumPy 2 (brainets/frites, brainets/frites)
Support mne-python versions newer than 1.5.0, up to the current release (brainets/frites)
Restore a compatibility shim so mne keeps working with NumPy >= 2.4, which removed np.in1d while mne still relies on it internally (brainets/frites)
Fix type comparison in frites.dataset.ds_utils.multi_to_uni_conditions() and in frites.plot.plot_conn_circle() (brainets/frites)

v0.4.4#

New Features#

Functions frites.conn.conn_reshape_undirected() and frites.conn.conn_reshape_directed() now supports multidimensional arrays (brainets/frites)
New function frites.core.ent_nd_g() to compute entropy on tensors (brainets/frites)
New function frites.conn.conn_ii() to estimate the interaction information (brainets/frites)
New function frites.conn.conn_pid() to estimate the partial information decomposition (brainets/frites)
New function frites.conn.conn_fit() to estimate the feature specific information transfer (brainets/frites#59) - aopy

Bug fixes#

Base 2 frites.core.ent_nd_g() (brainets/frites)

v0.4.3#

New Features#

frites.conn.conn_links() now accepts to use brain region names as source or target seeds (brainets/frites)
New function frites.stats.confidence_interval() for computing confidence intervals, standard deviation (sd) and standard error on the mean (sem) on numpy arrays and xarray DataArray (brainets/frites)
New plotting module with two functions for plotting connectivity matrices, frites.plot.plot_conn_heatmap() and frites.plot.plot_conn_circle() (brainets/frites, brainets/frites, brainets/frites, brainets/frites)
Add the possibility to control the node’s size in frites.plot.plot_conn_circle() (brainets/frites)
Add support for signed connectivity and improve node control of frites.plot.plot_conn_circle() (brainets/frites)

Bug fixes#

Fix parallel computing of frites.conn.conn_covgc() (brainets/frites)
Use linearly spaced indices instead of closest time points in frites.conn.define_windows() (brainets/frites)
frites.conn.conn_spec() average over tapers after computing connectivity metric (brainets/frites).Thanks to adam2392 and ruuskas from the mne-connectivity package.

v0.4.2#

New Features#

New function frites.simulations.sim_ground_truth() for simulating spatio-temporal ground-truths (brainets/frites)
New function frites.conn.conn_spec() for computing the single-trial spectral connectivity (brainets/frites) - ViniciusLima94
New method frites.workflow.WfMi.confidence_interval method to estimate the confidence interval (brainets/frites, brainets/frites, brainets/frites, brainets/frites)
New function frites.conn.conn_net() for computing the net connectivity (brainets/frites)
New function frites.set_mpl_style() for example styles
New function frites.conn.conn_links() for generating connectivity links (brainets/frites)
New function frites.utils.downsample() for down-sampling DataArray (brainets/frites)
frites.estimator.CorrEstimator can no be defined with Pearson or Spearman correlation with both vector or tensor-based implementations (brainets/frites)
frites.workflow.WfStats now allows to pass rfx_center parameter for controlling whether effect-sizes should be centered and rfx_sigma for the hat correction (brainets/frites)
frites.conn.conn_links() can now be used for selecting intra and / or inter-hemispheric connections (brainets/frites)
frites.conn.conn_links() allows selecting links with inter / intra / both roi connections (brainets/frites)

Internal changes#

Do not allow anymore to add new keys to the CONFIG dict (brainets/frites)

v0.4.1#

New Features#

New frites.estimator.CustomEstimator for defining custom estimators (brainets/frites, brainets/frites)
New function frites.conn.conn_fcd_corr() for computing the temporal correlation across networks (brainets/frites)
New function frites.utils.acf() for computing the auto-correlation (brainets/frites)
New function frites.conn.conn_ccf() for computing the cross-correlation (brainets/frites)

Bug fixes#

Fix attribute conversion in connectivity functions (brainets/frites)

v0.4.0#

New Features#

New estimators (frites.estimator.CorrEstimator, frites.estimator.DcorrEstimator) for continuous / continuous relationships (brainets/frites, brainets/frites, brainets/frites)
frites.conn.conn_dfc() supports passing other estimators (brainets/frites)
frites.utils.time_to_sample() and frites.utils.get_closest_sample() conversion functions (brainets/frites)
frites.conn.conn_ravel_directed() reshaping function (brainets/frites)
New frites.workflow.WfMi.copy for internal workflow copy (brainets/frites, brainets/frites)
New frites.workflow.WfMiCombine and example class for combining workflows (brainets/frites)
New frites.estimator.ResamplingEstimator trial-resampling estimator (brainets/frites)

Bug fixes#

Fix frites.workflow.WfMi.get_params when returning a single output (brainets/frites)
Improve attributes conversion for saving netcdf files (bool and dict) (brainets/frites, brainets/frites)
Fix Numpy np.float and np.int warnings related (brainets/frites, brainets/frites, brainets/frites)

v0.3.9#

New Features#

frites.conn.conn_dfc() supports multivariate data + improve computing efficiency (brainets/frites, brainets/frites)
Reshaping connectivity arrays support elements on the diagonal + internal drop of duplicated elements (brainets/frites)
frites.conn.conn_dfc() supports better channel aggregation (brainets/frites)

Internal Changes#

Connectivity metric now use the frites.dataset.SubjectEphy for internal conversion of the input data
frites.workflow.WfMi.get_params returns single-subject MI and permutations with dimension name ‘subject’ (instead of subjects) (brainets/frites)
All connectivity metrics now use frites.conn.conn_io() to convert inputs into a similar format
Improve CI

Bug fixes#

Fix frites.dataset.SubjectEphy when the data contains a single time point (brainets/frites)
Fix attributes of frites.conn.conn_covgc() (brainets/frites)
Fix frites.dataset.DatasetEphy representation without data copy + html representation (brainets/frites, brainets/frites#16)
Fix passing tail input to the frites.workflow.WfMi (brainets/frites)

v0.3.8#

New Features#

new frites.io.Attributes class for managing and printing datasets’ and workflow’s attributes (brainets/frites)
new frites.dataset.SubjectEphy single-subject container (brainets/frites)
new estimators of mutual-information, frites.estimator.GCMIEstimator (brainets/frites, brainets/frites, brainets/frites, brainets/frites), frites.estimator.BinMIEstimator (brainets/frites)
new kernel smoothing function frites.utils.kernel_smoothing()

Internal Changes#

Removed files (brainets/frites, brainets/frites, brainets/frites)
frites.dataset.DatasetEphy don’t perform internal data copy when getting the data in a specific ROI (brainets/frites)
Compatibility of MI estimators with workflows (brainets/frites, brainets/frites)
Improve the way to manage pairs of brain regions (brainets/frites, brainets/frites, brainets/frites, brainets/frites)

Breaking changes#

frites.dataset.SubjectEphy and frites.dataset.DatasetEphy to specify whether channels should be aggregated (default agg_ch=True) or not (agg=False) when computing MI. The agg_ch replace sub_roi (brainets/frites)
The workflow WfComod has been renamed frites.workflow.WfConnComod (brainets/frites)

Bug fixes#

Bug fixing according to the new version of frites.dataset.DatasetEphy (brainets/frites, brainets/frites, brainets/frites, brainets/frites, brainets/frites, brainets/frites, brainets/frites, brainets/frites)

v0.3.6#

New Features#

frites.dataset.DatasetEphy support multi-level anatomical informations (brainets/frites)

Internal Changes#

Connectivity functions have a better support of Xarray inputs (brainets/frites, brainets/frites, brainets/frites, brainets/frites)
Replace every string comparison ‘is’ with ‘==’ to test the content (brainets/frites)

v0.3.5#

New Features#

New function for reshaping undirected connectivity arrays (like DFC) frites.conn.conn_reshape_undirected() (brainets/frites, brainets/frites)
New function frites.utils.savgol_filter() that works on DataArray (brainets/frites)
New function for reshaping directed connectivity arrays (like COVGC) frites.conn.conn_reshape_directed() (brainets/frites)
New method frites.workflow.WfMi.get_params in order to get the internal arrays formatted as DataArray (brainets/frites)
Integration of MNE’s progress bar (brainets/frites, brainets/frites, brainets/frites)
Possibility to cache computations in the parallel function

Internal Changes#

DataArray now contain a name such as a type to make it clear what is it (brainets/frites)
sigma parameter when performing the t-test can be changed though the CONFIG file (brainets/frites)

Bug fixes#

Fix ttested attribute when saving (brainets/frites)
Fix computing the sigma across all ROI since it uses the maximum over every axes (brainets/frites)
Fix high RAM consumption when computing the pop_mean_surr (brainets/frites)

v0.3.4#

Bug fixes#

Fix frites.workflow.WfMi.conjunction_analysis for seeg data (brainets/frites)

Breaking changes#

Xarray is now the default output type as it supports the definition of multi-dimensional containers with label to each coordinates (brainets/frites) + illustrating examples of how to use xarray (brainets/frites)
Every connectivity measures have been moved to frites.conn (brainets/frites)
Deep support for correcting for multiple-comparisons using multi-dimensional data (brainets/frites, brainets/frites) + support for frites.workflow.WfStatsEphy (brainets/frites) + support for frites.dataset.DatasetEphy (brainets/frites) + support for frites.workflow.WfMi (brainets/frites)

New Features#

New class frites.simulations.StimSpecAR for generating Auto-Regressive Models (brainets/frites, brainets/frites)
Conditional Covariance based Granger Causality + example (brainets/frites)

Internal Changes#

Improve testings (brainets/frites)

v0.3.3#

Internal Changes#

frites.workflow.WfFit and frites.workflow.WfConn are now using frites.dataset.DatasetEphy.get_connectivity_pairs (brainets/frites)
Improve warning messages + assert error for negative determinant for frites.core.covgc

New Features#

New method frites.dataset.DatasetEphy.get_connectivity_pairs in order to get possible connectivity pairs (brainets/frites)
New function frites.utils.define_windows() and frites.utils.plot_windows() in order to generate and plot slicing windows + tests + example (brainets/frites)
New function for computing the DFC frites.core.dfc_gc (brainets/frites)
When using DataArray with frites.core.dfc_gc and frites.core.covgc, temporal attributes are added (brainets/frites, brainets/frites)
New function for computing the covgc frites.core.covgc (brainets/frites)
Step parameter for frites.core.covgc (brainets/frites)
frites.core.covgc can no be computed using Gaussian-Copula (brainets/frites)
Add frites.workflow.WfMi.conjunction_analysis for performing conjunction analysis + example (brainets/frites)

Bug fixes#

Fix when data contains a single time point (brainets/frites)
Fix mi model and mixture (1d only) (brainets/frites)

v0.3.2#

Breaking changes#

Avoid duplicates dataset construction when using MNE / xarray (brainets/frites, brainets/frites)
frites.dataset.DatasetEphy supports None for the y input (brainets/frites)

Internal Changes#

Dtypes of y and z inputs are systematically check in frites.dataset.DatasetEphy in order to define which MI can then be computed (brainets/frites)

New Features#

frites.dataset.DatasetEphy supports Xarray inputs + selection though coordinates (brainets/frites)
New workflow for computing pairwise connectivity frites.workflow.WfConn (brainets/frites)

Documentation#

Adding new examples for creating datasets (brainets/frites)

v0.3.1#

Breaking changes#

change frites.workflow.WfFit input directed for net (brainets/frites, brainets/frites#1)
The GCRN is automatically defined (per subject when RFX / across subjects when FFX) (brainets/frites)
Remove the level input parameter + only mcp is used + only maxstat when performing cluster based (brainets/frites)

New Features#

frites.dataset.DatasetEphy support spatio-temporal slicing (brainets/frites), resampling (brainets/frites) and savitzki-golay filter (brainets/frites)
Support setting random_state in frites.workflow.WfMi and frites.workflow.WfFit (brainets/frites)
DataArray outputs contains attributes that reflect the configuration of the workflow (brainets/frites)

Bug fixes#

Fix multi-sites concatenation (brainets/frites) in frites.dataset.DatasetEphy
Fix p-values to zeros in frites.workflow.WfFit (brainets/frites)
Fix FIT outputs for 3D arrays and DataArray (brainets/frites)

Internal Changes#

Remap multiple conditions when integers (brainets/frites) in frites.dataset.DatasetEphy
Workflows now have an internal configuration (brainets/frites)

Documentation#

Reformat examples gallery (brainets/frites)