Skip to content

Latest commit

 

History

History
423 lines (335 loc) · 28.1 KB

changes.md

File metadata and controls

423 lines (335 loc) · 28.1 KB
Raw

What's New

0.4.0

🛠 Breaking Changes

  • Change default highpass filter cutoff from 0.009 to 0.01 by @kahinimehta in #755
  • Refactor anatomical workflow by @tsalo in #706
  • Do not require --combineruns to generate DCAN QC files by @tsalo in #753
  • Change QC filenames and fix xcp_d-combineqc command-line interface by @tsalo in #762
  • Retain cohort entity in derivative filenames by @tsalo in #769
  • Require the --cifti flag in order to use --warp-surfaces-native2std by @tsalo in #770
  • Zero out parcels with <50% coverage by @tsalo in #757
  • Use constant padding and maximum padlen for temporal filtering by @tsalo in #779
  • Replace non-aggressive denoising with orthogonalization and streamline denoising by @tsalo in #808
  • Add design matrix to report by @tsalo in #824
  • Write out censored results by @tsalo in #820
  • Allow users to disable censoring, and only generate ALFF if censoring is disabled by @tsalo in #828
  • Implement --min-time parameter by @tsalo in #836
  • Remove deprecated --dummytime parameter by @tsalo in #837
  • Change FD Threshold from 0.2 to 0.3 by @kahinimehta in #854
  • Use filtered motion in nuisance regression by @tsalo in #871

🎉 Exciting New Features

  • Add option to estimate brain radius from brain mask by @tsalo in #754
  • Output warped atlases to derivatives by @tsalo in #647
  • Add min_coverage parameter to threshold parcel coverage by @tsalo in #782
  • Improve executive summary carpet plots by @tsalo in #747
  • Output T2w images if available by @tsalo in #648
  • Generate CIFTI and TSV versions of coverage, timeseries, and correlation files by @tsalo in #785
  • Add colorbar to executive summary carpet plots by @tsalo in #829
  • Support fsLR-space shape files generated by preprocessing pipelines by @tsalo in #773
  • Support preprocessing derivatives with T2w, but no T1w by @tsalo in #838
  • Support high-pass or low-pass only filtering by @smeisler in #862

🐛 Bug Fixes

  • Replace missing vertices' values with NaNs by @tsalo in #743
  • Select MNI152NLin6Asym target space for T1w from CIFTI derivatives by @tsalo in #759
  • Only generate brainsprite figures when the --dcan_qc flag is used by @tsalo in #766
  • Check for existence of dataset_description.json in fmri_dir by @tsalo in #806
  • Move dataset desc check to after conversion by @tsalo in #809
  • Refactor HCP/DCAN ingression and fix converted filenames by @tsalo in #714
  • Connect custom confounds to confound consolidation node by @tsalo in #835
  • Work around load_confounds aCompCor bug by @tsalo in #851
  • Allow for smoothing to be zero by @tsalo in #861
  • Use appropriate T1w/T2w in brainsprite workflow by @tsalo in #860
  • Test DCAN and HCP ingestion and fix related bugs by @tsalo in #848

Other Changes

  • Add tests for cifti smoothness by @kahinimehta in #659
  • Fix CIFTI downcasting test by @tsalo in #752
  • Use pytest to test command-line interface by @tsalo in #740
  • Upload coverage reports to CodeCov by @tsalo in #758
  • Link to xcp_d-examples in documentation by @tsalo in #761
  • Add information about preprocessing requirements by @tsalo in #772
  • Replace MultiLabel interpolation with GenericLabel by @tsalo in #786
  • Mention cosine regressors in aCompCor boilerplate by @tsalo in #788
  • Update base Docker image by @tsalo in #799
  • Document Singularity requirements by @tsalo in #805
  • Use workflow nodes for workflow tests by @tsalo in #807
  • Adjust code for niworkflows 1.7.3 by @tsalo in #810
  • Use a workflow for concatenation by @tsalo in #821
  • Create prepare_confounds and denoise_bold workflows by @tsalo in #827
  • Simplify the anatomical workflow some more by @tsalo in #841
  • Remove codecov dependency by @tsalo in #850
  • Add boilerplate to the executive summary by @tsalo in #857
  • Add additional argument aliases by @smeisler in #870
  • Update docs to mention HCPYA version by @kahinimehta in #869
  • Improve boilerplate by @tsalo in #866
  • Update gitignore by @tsalo in #874
  • Fix issues with HCP Ingression by @kahinimehta in #872
  • Document how to process multiple, but not all, tasks by @tsalo in #876
  • Describe scope of XCP-D by @tsalo in #878
  • Update docs by @kahinimehta in #882
  • Always generate the DCAN executive summary by @tsalo in #888
  • Clean up testing framework and documentation by @tsalo in #889

Full Changelog: https://github.com/PennLINC/xcp_d/compare/0.3.2...04.0

0.3.2

This release adopts a workbench show-scene-based brainsprite for the executive summary. It also removes the brainsprite figure from the nipreps report.

🎉 Exciting New Features

  • Adopt executive summary's brainsprite using jinja templates by @tsalo in #702

Other Changes

  • Remove example data and fix workflow graphs by @tsalo in #738
  • Replace CiftiDespike with connected nodes by @tsalo in #737

Full Changelog: https://github.com/PennLINC/xcp_d/compare/0.3.1...0.3.2

0.3.1

🛠 Breaking Changes

  • Remove --nuissance-regressors and --bandpass_filter by @tsalo in #658

🎉 Exciting New Features

  • Add --dummy-scans parameter and deprecate --dummytime by @tsalo in #616
  • Add --bids-filter-file parameter by @tsalo in #686
  • Enable non-aggressive denoising with signal regressors by @tsalo in #697
  • Improve identification of FreeSurfer derivatives by @tsalo in #719
  • Collect preprocessed surfaces in new function by @tsalo in #731

🐛 Bug Fixes

  • Remove dummy volumes from custom confounds files by @tsalo in #660
  • Remove dummy volumes from beginning of each run in concatenation workflow by @tsalo in #661
  • Retain noise components instead of dropping them in load_aroma by @tsalo in #670
  • Use input_type to determine order of preferred spaces by @tsalo in #688
  • Infer volumetric space from transform in executive summary with cifti data by @tsalo in #689
  • Change input type from HPC to HCP by @kahinimehta in #691
  • Replace failing non-aggressive AROMA denoising with working aggressive denoising by @tsalo in #693
  • Select best volumetric space for transforms even when using --cifti flag by @tsalo in #695
  • Downcast >32-bit files to 32-bit by @tsalo in #666
  • Use appropriate intent codes for cifti outputs by @tsalo in #690
  • Only concatenate processed runs by @tsalo in #713
  • Do not use downcasted files as name sources by @tsalo in #712
  • Warp segmentation file to appropriate space for carpet plots by @tsalo in #727
  • Use brain mask in NIFTI connectivity workflow by @tsalo in #733

Other Changes

  • Consolidate confounds at beginning of denoising workflows by @tsalo in #664
  • Remove unused outputnodes from nifti and cifti workflows by @tsalo in #667
  • Move QC/censoring plots into new workflow by @tsalo in #668
  • Testing affines don't change across XCP runs by @kahinimehta in #649
  • Load confounds via Nilearn functionality by @kahinimehta in #675
  • Update docs by @kahinimehta in #682
  • Add masks to package data by @kahinimehta in #715
  • Use ds001491 fMRIPrep derivatives for tests by @tsalo in #698
  • Add tests for confound loading function by @tsalo in #730
  • Refactor executive summary workflow by @tsalo in #721
  • Track start of workflow with sentry by @tsalo in #732

Full Changelog: https://github.com/PennLINC/xcp_d/compare/0.3.0...0.3.1

0.3.0

We are happy to announce a new minor release, with several backwards-incompatible changes.

Two big breaking changes are (1) there is a new --dcan-qc flag that determines if the executive summary and DCAN-format QC files will be generated, and (2) custom confounds should now have headers, should be tab-delimited, and should have the same names as the fMRIPrep confounds, for easier indexing.

🛠 Breaking Changes

  • Output ReHo as a CIFTI by @tsalo in #601
  • Add --dcan-qc flag by @tsalo in #650
  • Support custom confounds with headers by @tsalo in #642

🎉 Exciting New Features

  • Use BIDSLayout in concatenation code by @tsalo in #600
  • [ENH] Support Nibabies ingression into XCP by @kahinimehta in #602
  • Add static ALFF and ReHo plots to report by @kahinimehta in #604
  • Write confounds out to derivatives by @kahinimehta in #634
  • Add column names to confounds df by @kahinimehta in #641

🐛 Bug Fixes

  • Select BOLD files in a single space by @tsalo in #603
  • Censor data in executive summary plots correctly by @kahinimehta in #614
  • Only generate ALFF derivatives if bandpass filtering is enabled by @tsalo in #628
  • Do not merge in concatenated files when re-running concatenation workflow by @tsalo in #633
  • Map abbreviated entities to full names for templateflow get call by @tsalo in #654
  • Ensure dummyscans is an integer in PlotSVGData by @tsalo in #655

Other Changes

  • Use Nilearn for brainsprite generation by @tsalo in #607
  • Cache the downloaded test data by @tsalo in #629
  • Use BIDSLayout instead of globbing functions to collect necessary files by @tsalo in #621
  • Start to standardize interface parameter calls by @tsalo in #638
  • Simplify transform-getting functions by @tsalo in #623
  • Lint with black without linting workflow connections by @tsalo in #640

Full Changelog: https://github.com/PennLINC/xcp_d/compare/0.2.0...0.3.0

0.2.2

This is a patch release for the 0.2 series. The main bug being fixed is that using --dummytime was causing crashes in the executive summary workflow.

🎉 Exciting New Features

  • Write confounds out to derivatives by @kahinimehta in #634
  • Add column names to confounds df by @kahinimehta in #641

🐛 Bug Fixes

  • Do not merge in concatenated files when re-running concatenation workflow by @tsalo in #633
  • Map abbreviated entities to full names for templateflow get call by @tsalo in #654
  • Ensure dummyscans is an integer in PlotSVGData by @tsalo in #655

Other Changes

  • Cache the downloaded test data by @tsalo in #629
  • Use BIDSLayout instead of globbing functions to collect necessary files by @tsalo in #621
  • Start to standardize interface parameter calls by @tsalo in #638
  • Simplify transform-getting functions by @tsalo in #623
  • Lint with black without linting workflow connections by @tsalo in #640

Full Changelog: https://github.com/PennLINC/xcp_d/compare/0.2.1...0.2.2

0.2.1

This is a patch release for 0.2.0.

There is a known bug with the concatenation workflow, so we advise users not to use the -m/--combineruns option with this release.

🛠 Breaking Changes

  • Output ReHo as a CIFTI by @tsalo in #601

🎉 Exciting New Features

  • Use BIDSLayout in concatenation code by @tsalo in #600
  • [ENH] Support Nibabies ingression into XCP by @kahinimehta in #602
  • Add static ALFF and ReHo plots to report by @kahinimehta in #604

🐛 Bug Fixes

  • Select BOLD files in a single space by @tsalo in #603
  • Censor data in executive summary plots correctly by @kahinimehta in #614
  • Only generate ALFF derivatives if bandpass filtering is enabled by @tsalo in #628

Other Changes

  • Use Nilearn for brainsprite generation by @tsalo in #607

Full Changelog: https://github.com/PennLINC/xcp_d/compare/0.2.0...0.2.1

0.2.0

This is a big release. There are a lot of backwards-incompatible changes, as well as a number of bug-fixes and enhancements.

There is a full list of the changes made between 0.1.3 and 0.2.0 below. However, here are some highlights:

  1. We have renamed and reorganized a number of the outputs created by xcp_d to better follow BIDS convention. There is a lot in xcp_d that falls outside the current BIDS specification, so we took inspiration from a number of BIDS Extension Proposals (BEPs) that folks have written over the years.
  2. There is a new --warp-surfaces-native2std flag, which produces a number of subject-specific surfaces in fsLR space. This was previously run by default in version 0.1.3. The workflow that this flag triggers is also much improved, thanks to @madisoth.
  3. We have fixed a major bug, in which the parameter set users selected was ignored, and "36P" was used no matter what.
  4. The HTML reports have been improved. The interactive segmentation image from the executive summary has been added to the main report, along with the BOLD-T1w coregistration figure. We have also added a new plot to show the impact of filtering motion parameters on censoring.

🛠 Breaking Changes

  • Ensure TSV files are tab-delimited by @tsalo in #541
  • Rename derivatives to be more BIDS-compliant by @tsalo in #553
  • Replace --func-only with --warp-surfaces-native2std by @tsalo in #562
  • Add headers to motion.tsv and outliers.tsv files by @tsalo in #587

🎉 Exciting New Features

  • Create dataset description file by @tsalo in #561
  • Distinguish preprocessed dataset formats by @tsalo in #567
  • Output temporal mask by @tsalo in #586
  • Add censoring plot to summary reports by @tsalo in #579
  • Add existing SVG figures and brainplot HTML figure to HTML report by @tsalo in #590
  • Output all filtered motion parameters by @tsalo in #592
  • Deprecate --bandpass_filter in favor of --disable-bandpass-filter by @tsalo in #588

👎 Deprecations

  • Deprecate nuissance-regressors in favor of nuisance-regressors by @tsalo in #513

🐛 Bug Fixes

  • Fix fsLR32k reg and ApplyWarpfield issues by @madisoth in #442
  • Replace MIT license with BSD-3 by @tsalo in #457
  • Update utils.py to fix MNILin6 imports by @smeisler in #458
  • Potential fix for "no tasks found" problem by @kahinimehta in #464
  • Fix "manadatory" typo in interfaces by @tsalo in #540
  • Change "hpc" in CLI to "hcp" by @tsalo in #560
  • Pin pybids to version 0.15.1 by @madisoth in #537
  • Fix resampling of fsnative structural surfaces to fsLR by @madisoth in #496
  • Fix documentation and validation of motion filtering parameters by @tsalo in #575
  • Show framewise displacement before filtering in QCPlot by @tsalo in #581
  • Fix concatenation code w.r.t. recent output changes by @tsalo in #593
  • Use user-requested parameters in regression by @tsalo in #596
  • Update load_aroma so the correct file is being read in by @kahinimehta in #599

Other Changes

  • Rf/smoothing by @kahinimehta in #415
  • Rf/misc. by @kahinimehta in #421
  • Rf/fconts by @kahinimehta in #424
  • [RF] Miscellaneous by @kahinimehta in #439
  • [RF] Reho Computation by @kahinimehta in #437
  • [DOC] Update docs by @kahinimehta in #433
  • [RF] ALFF Computation by @kahinimehta in #435
  • [RF] fixes by @kahinimehta in #443
  • [RF] Executive summary by @kahinimehta in #445
  • [RF] Compute qcplot.py by @kahinimehta in #447
  • [RF] Make TR tests shorter by @kahinimehta in #462
  • [DOC] Add badges to README by @tsalo in #466
  • [DOC] Add issue templates and route to NeuroStars by @tsalo in #465
  • [REF] Use nilearn for loading and masking nifti data by @tsalo in #459
  • [TST] Add linting GH workflow by @tsalo in #467
  • [DOC] Address formatting issues in documentation by @tsalo in #469
  • [REF] Replace all relative imports with absolute ones by @tsalo in #472
  • [FIX] Address failing RTD build by @tsalo in #477
  • [REF] Use f-strings consistently throughout codebase by @tsalo in #473
  • [DOC] Use welcome bot to comment on new contributors' PRs by @tsalo in #475
  • [RF] Linting by @kahinimehta in #479
  • [DOC] Add or format docstrings throughout package by @tsalo in #474
  • [REF] Lint interfaces by @tsalo in #480
  • [REF] Lint workflows by @tsalo in #481
  • [REF] Lint utils by @tsalo in #485
  • [REF] Rename functions to snake_case by @tsalo in #487
  • [REF] Rename classes to CamelCase by @tsalo in #486
  • [RF] Get rid of unused variables by @kahinimehta in #493
  • [DOC] Add docstrings to CLI module by @tsalo in #490
  • [REF] Address remaining linter warnings by @tsalo in #494
  • [REF] Move functions and classes into appropriate modules by @tsalo in #489
  • [DOC] Add API to documentation by @tsalo in #498
  • [DOC] Clean up workflow docstrings by @tsalo in #503
  • [REF] Remove duplicate functions by @tsalo in #502
  • [REF] Empty init.py files by @tsalo in #499
  • [REF] Centralize DerivativesDataSink definition by @tsalo in #510
  • [MAINT] Add CITATION.cff file by @tsalo in #482
  • [TST] Store CircleCI artifacts by @tsalo in #521
  • [REF] Start replacing common parameters with fill_doc by @tsalo in #509
  • [REF] Remove the notebooks module by @tsalo in #531
  • [DOC] Add TemplateFlow citation to boilerplate by @tsalo in #520
  • [DOC] Use sphinxcontrib-bibtex to embed BibTeX references in documentation by @tsalo in #516
  • [TST] Make local tests stricter by @tsalo in #545
  • [TST] Check for existence of outputs in tests by @tsalo in #542
  • [REF] Remove FSL from dependencies by @tsalo in #528
  • [TST] Restructure local patch handling in testing setup by @tsalo in #550
  • [REF] Simplify BOLD/CIFTI post-processing workflow call by @tsalo in #534
  • [REF] Remove unused init_post_process_wf by @tsalo in #535
  • [REF] Use MapNodes to simplify functional connectivity workflows by @tsalo in #546
  • Add release notes template by @tsalo in #552
  • Add warning to documentation about M1 chip by @tsalo in #572
  • Improve confound plot readability and fix moving registration plot by @kahinimehta in #530
  • Fix the local pytest script and documentation by @tsalo in #564
  • [TEST] Add tests for outstanding modules by @kahinimehta in #529
  • [FIX] Fcon workflow tests are incompatible with changes from main by @kahinimehta in #584

New Contributors

  • @tsalo made their first contribution in #457

Full Changelog: https://github.com/PennLINC/xcp_d/compare/0.1.3...0.2.0

0.1.3

  • Add analysis-level option
  • Quick fix for bugs in bold and cifti workflow

Full Changelog: https://github.com/PennLINC/xcp_d/compare/0.1.2...0.1.3

0.1.2

  • Fix malloc failure by @kahinimehta in #395
  • [RF] Filtering by @kahinimehta in #390
  • Rf/rename by @kahinimehta in #398
  • Rf/rename by @kahinimehta in #400
  • Rf/rename by @kahinimehta in #401
  • [RF] Smoothing by @kahinimehta in #404

Full Changelog: https://github.com/PennLINC/xcp_d/compare/0.1.1...0.1.2

0.1.1

  • Replace NiPype's filemanip.py by @madisoth in #383
  • [RF] Fix typos by @kahinimehta in #386
  • ENH: Update filemanip.py "special_extensions" with the rest of the CIFTI-2 multi-part extensions by @madisoth in #388
  • [RF] Interpolation by @kahinimehta in #387
  • FIX: reho / write_gii by @madisoth in #385

Full Changelog: https://github.com/PennLINC/xcp_d/compare/0.1.0...0.1.1

0.1.0

Additional features for surface processing and expanded CI testing

0.0.9

  1. Now supports Schaefer 100,200,300,400,500,600,700,800,900,1000
  2. CompCor fix, where previously we were not including cosine
  3. Removed Temporal CompCor; no one should use this anyways
  4. Added aroma_gsr and compcor_gsr, which include global signal regression
  5. Huge update of the documentation
  6. This is the handoff update from @a3sha2 to @mb3152

0.0.8

afni-reho-despike correction

  • Hcpdcan by @a3sha2 in #211
  • Hcpdcan by @a3sha2 in #212

Full Changelog: https://github.com/PennLINC/xcp_abcd/compare/0.0.7...0.0.8

0.0.7

  • Hcpdcan by @a3sha2 in #181
  • Hcpdcan by @a3sha2 in #182
  • Hcpdcan by @a3sha2 in #183
  • thread/meg_gb included in plugin settings by @a3sha2 in #184
  • nthread/ompthread by @a3sha2 in #185
  • remove labelled by @a3sha2 in #186
  • increased bold memgb size by @a3sha2 in #187
  • Hcpdcan by @a3sha2 in #194

Full Changelog: https://github.com/PennLINC/xcp_abcd/compare/0.0.6...0.0.7

0.0.6

anatomical update executive summary dcan ingression

0.0.5

NIBABIES TEST

  • MNI152NLin6Asym template for nibabies #156
  • update executive summary and report #69 #115 #157 #151
  • anatomical workflow added including freesurfer/freesurfer*
  • outputs for cifti and nifti

0.0.4

final release for hcp

0.0.3

hcp Update cifti.py

0.0.2

Merge pull request #113 from PennLINC/template

Template

0.0.1

Merge pull request #64 from PennLINC/test_xcp

add version automatically