Commit
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
pypi-nibabel: Autospec creation for update from version 1 to version …
…3.2.1
Alexandre Gramfort (13):
ENH : adding support for some FreeSurfer files
STY: missing ending line
ENH : adding freesurfer to setup.py
ENH : adding nibabel.freesurfer.tests to setup.py
ENH: using nose.assert_true
ENH : removing freesurfer.io.read_scalar_data now that MGH format is properly handled
FIX : fix the read_annot when parcellation has missing values and orig_ids=False.
FIX : handle versions of fsaverage
FIX : dtype pb with annot on numpy 1.5.1
add test + misc
address comments
address comments
don't count warnings
Anibal Sólon (11):
make sure cov is installed
pytest port
drop the nose
finish porting gifti module tests to pytest
make sure cov is installed
test script condition
reintroduce nosetests into matrix
remove unnecessary dependency
pytest is already tested
nose test inclusions
windows breaklines..?
Ariel Rokem (27):
BF: Use newer call to frombytes, replacing fromstring.
PEP8
Add pillow to dependencies, add old pillow to test with old pydicom.
Test with PIL on old Python.
Compatibility with older versions of PIL that don't have frombytes.
Revert back to older pillow, since PIL is a pill.
TST: If a file doesn't exist, we want an informative error.
RF: Add testing for file existence up front.
TST: Use IOError for Python 2 support.
Use FileNotFoundError. Implement in py3k for Python 2 compatibility.
PEP8
TST: Add testing on Python 3.5 on Appveyor.
Explicitly create a conda environment for the test, specify Python version.
Separate things out a bit, set always-yes for conda.
Test on Python 3.5 with 32 architecture, to see if that matters.
TST: Make sure that img.__getitem__ exception throws the right error.
RF: Usually indexing is done with an indexing item.
TST: Add explanatory comment.
RF: Make idx a positional argument.
Fix in coordinate_systems tutorial
RF: Circumvents a deprecation warning from `np.fromstring`
RF: fromstring => frombuffer in externals.netcdf.
RF: fromstring => frombuffer in gifti/cifti2
RF: fromstring => frombuffer.
BF: Use a writeable bytearray for the header data.
NF: Implement a `readinto` method for Opener.
Don't refcheck on this resize, to avoid errors
B. Nolan Nichols (7):
updated image_orient_patient function
few updates to naming conventions for edicom sequences - e.g., PixelMeasuresSequence
instantiate superclass in MultiframeWrapper __init__
added is_edicom flag
fixed sop class comparison
PL: updated docstring quotes and other recommended changes
RF+TST: updated multiframe error handling and tests
Bago Amirbekian (4):
BF - fix RGB read issue + test
TEST - set imdtype with code+npt
TEST - added test for set_qform and set_sform
BF - only check dtype min-max is clip is needed
Basile Pinsard (5):
fix decimal rescale values causing Decimal data frame
adding comment about float casting
fix endian value in gifti
fix b64encoding for no newline in outputs
fixes + backcompat test
Ben Cipollini (144):
Add an 'axis' parameter to concat_images, plus two tests.
Try again, this time with lists and more tests...
Add greater coverage of different shapes.
Make this work for all 3D and 4D combinations possible, across all axes possible. Test extensively.
Allow mixed files and objects.
Improve efficiency: load img0 once, del reference, and don't check affine on first image.
Add a final comment.
Code reorganization, comment editing.
Remove 3D/4D special-case code.
Fix broadcasting bug (img1.shape == (1,2,3), img2.shape=(1,2,1), axis=None would broadcast rather than error)
Similar bug in axis=int pathway, due to np.concatenate "smartness".
Test 2D - 5D; remove some tests to increase speed.
Convert exceptions to string.
Remove default argument.
Small code review tweaks.
Fix for #324 - use consistent syntax for have_scipy.
Change idioms for selecting the 2nd output arg.
Simplify idioms, delete unnecessary 'global' definition.
Linting nifti1.py
Allow classes to register extensions with opener functions; use this for .mgz. Remove .mgz-specific logic from BinOpener, and deprecate.
Use Opener (which contains .mgz) instead of BinOpener (which is no longer needed).
Move decorator to ImageOpener, use ImageOpener.
register_extension => register_ext_from_image
Add simple tests for adding a new file association.
Add test for BinOpener deprecation.
Tweak docstrings / parameter name.
New interface: suppress_warnings, clear_and_catch_warnings, and error_warnings.
Unlike numpy, default to record=True (this is what we had before).
Add deprecation tests
Fixing test failures.
Deprecate the use of class_map and ext_map.
Migrating imageclasses, away from IMAGE_MAP
Improve error handling, efficiency, and search more broadly over header extensions.
Modify save, remove vestigates of class_map / ext_map
Code cleanup after self code review, fix for Python 3 'filter' issue.
Remove references to BinOpener
Linting, cleaning up logic.
Adding tests for loading each image type.
Add back analyze header tests.
sniff_size => sizeof_hdr
Simplify is_image test.
RF: getArraysFromIntent => get_arrays_from_intent
RF: make meta, numDA, and labeltable into properties.
TST: Add LabelTable test, test cleanup
Add get_/set_ function deprecations and deprecation tests.
get_metadata => metadata
get_rgba => rgba
TST: Add tests for darrays / numDA
BF: can't use assert_raises as context (py2.6)
Add tests for rgba, clean code.
Check NumberofDataArrays, warn only if it's wrong.
Add tests, change print error to ValueError, test it
ValueError=>TypeError
Tweak error message, add all TypeError tests, localize properly.
Make sure xml is encoded as utf-8
RF: convert xml code to use xml library.
Move to XmlSerializable interface.
RF: add thin layer around xml.etree.ElementTree and share XmlSerializable
RF: remove data_tag
deprecate gifti.data_tag
deprecate (not delete) to_xml_start, to_xml_end
RF: Create abstract FileBasedImage / FileBasedHeader classes.
RF: combine BaseHeader/Image with FileBasedHeader/Image.
RF: Header=>SpatialHeader
get_header => header; deprecate get_header
Fix filebasedimage docstring.
basestring => six.string_types
Remove any mention of affine or dataobj from FileBasedImage
make from_image an abstract method
Syncing up with code changes to spatialimages.py that occurred while I was doing this PR.
Make Gifti extend FileBasedImage, implement from_file_obj, to_file_obj.
Deprecate giftiio
STY: Ignore output of coverage --html
Make GIFTI call superclass __init__
Code cleanup for @effigies code review.
xmlutils.py => xmlbasedimages.py
Base GiftiImage on XmlBasedImage
Use gifti's to_file_map as XmlBasedImage's default implementation.
Create a new XmlImageParser abstract class, which must be implemented and attached to each XmlBasedImage subclass.
Implement GiftiImageParser, stamp on GiftiImage
Fix: GiftiImage should use get_/set_filename
Properly deprecate Outputter, parse_gifti_file
BF: re-implement buffer size; makes parsing faster!
TST: Add parse_gifti_fast deprecation tests (and fix old deprecation tests)
Use verbose=1 instead of global DEBUG_PRINT flag
STY: linting parse_gifti_fast
STY, DOC: linting, documentation.
Allow buffer_size=None, to avoid setting explicitly.
FIX: xmlbasedimages
Remove XmlBasedImage, xmlbasedimages.py=>xmlutils.py
Deprecate (not delete) Outputter.initialize.
Match previous interface for Outputter, add back minimal docstring.
DOC: Add six copyright (#375)
DOC: fix openhub.net badge (#378), prevent overflow of sidebar content.
DOC: change URL to numpy docstring style guide (partial fix for #376)
DOC: include all copyright information in greyed box.
BF: Fix #362 using buffered gzip read.
TST: add image tests in large image test.
TST: Add a runif_extra_has decorator.
DOC: updates for code review.
BF: Only define code change in Py3.5.0; only use if needed.
TST: Fix logic to check test values, use ones instead of zeros (zeros returned data read/copy fails).
STY: code changes per code review.
DOC: add documentation of advanced testing.
STY: autopep8 on source code.
STY: autopep8 on test code.
STY: autoflake on source code.
STY: autoflake on test files.
update tox
STY: flake8 manual fixes
BF: fix bug in mriutils use of __all__
STY: manual fixes to autopep8
TST: add flake8 of core nibabel code to TravisCI
TST: remove unnecesary checks, document necessary checks
BF: files must be makeable (they're made!) for this test.
save conversion logic is not as strong as needed (should really use maybe_is_image). Use some try...catch to safeguard.
Localize Cifti to /cifti; use nib.load/save
ENH: add xmlutils.XmlParser
Use _to_xml_element, snake_case properties, remove numXXX props, never return empty when xml requested.
Use assert_equal(a,b) instead of assert_true(a==b)
fix: move intent_codes.
fix: no xmlimage
STY: pep8 updates
Add a write_morph_data function.
Add a write_morph_data function.
cifti=>cifti2, Cifti=>Cifti2
FIX: Take care of byte vs. string / BytesIO vs. StringIO encodings for Python 3.
deprecation tests should actually test for warnings.
data arrays created via constructor are malformed; create via from_array.
Do not declare class variables as types; they mask failure to set instance-level variables.
make num_dim a read-only property
STY: fix up pep8 failures.
TST: add gifti string and MetaData tests.
Add a write_morph_data function.
BF: files must be makeable (they're made!) for this test.
save conversion logic is not as strong as needed (should really use maybe_is_image). Use some try...catch to safeguard.
Localize Cifti to /cifti; use nib.load/save
Use _to_xml_element, snake_case properties, remove numXXX props, never return empty when xml requested.
Use assert_equal(a,b) instead of assert_true(a==b)
fix: move intent_codes.
fix: no xmlimage
STY: pep8 updates
cifti=>cifti2, Cifti=>Cifti2
FIX: Take care of byte vs. string / BytesIO vs. StringIO encodings for Python 3.
Benjamin C Darwin (1):
test_proxy_api: replace nonexistent `extend` method call on a tuple with `+=`
Bennet Fauber (2):
set sform with set_sform not get_sform
two withs become one
Bertrand Thirion (1):
removed dependence on functools.reduce
Brendan Moloney (22):
BF+TST - Allow 0 len items in CSA hdr if VM != 0
BF: Only use the cross product (possibly negated) for the slice normal.
DOC: Added some comments to clarify what is going on.
TST: Make sure the slice normal is sufficiently orthogonal, even when CSA slice normal is not.
TST: Additional tests for getting slice normal from DicomWrapper
ENH: Added funtions to transform orientations.
DOC: Fixed typo in axcodes2ornt docstring
BF: ornt_transform was returning the inverse of the correct orientation.
TST: Added test for decimal rescale issue.
ENH: Handle 4D+ datasets.
ENH: Improve get_data for 4D+ multiframe
BF+TST: Fixed lookups for iop/ipp/pixspacing
TST: Add test data for previous commit
TST: Point multiframe test at correct file.
TST: Add test that makes sure resulting data order doesn't change for 4D multiframe.
CMP: Fix Python3 incompatability in last commit.
DOC: Added info about how blocks of private elements are reserved
DOC: Improve description of private block reservation.
BF+TST: Skip over missing CSA header elem instead of raising.
RF: Move dicom import into test function
RF: Skip test for python < 2.6 and pydicom < 1.0
ENH: Add parser for Siemens "ASCCONV" text format
Cameron Riddell (15):
load save with pathlib.Path objects
Add fn to coerce pathlike to string
Insert _stringify_path to entry points
Refactor _stringify_path to filename_parser.py
resolve flake8
testing for pathlib
purge basestring
move stringify guard to types_filenames
update docstrings to accept str or os.PathLike
update _stringify_path doc
update docstrings to accept str or os.PathLike
mghformat accept pathlib for filespec_to_file_map
tests pathlib compatible
update _stringify_path doc
Fix flake 8 issues
Carl Gauthier (1):
DOC: Fixed a typo
Christian Haselgrove (6):
Import the data format tools
dicomfs fixes regarding encoding and other bits.
removed explicit tags (in favor of dicom object attributes)
importing dft from nibabel
dft tests and test data
templating HTML output
Christopher J. Markiewicz (757):
Add writing capabilities for freesurfer triangle files
Use more modern tempfile methods, ensure tests run, improve writing
Add default timestamp behavior to write_geometry
NF: Write capabilities for freesurfer triangle files
TEST: Use portable getpass.getuser(); use InTemporaryDirectory
BF: Use portable getpass.getuser() in write_geometry
STY: pep8 clean
ENH Add read_scalars parameter to freesurfer.read_label() (used in PySurfer)
BF Add parameter to freesurfer.read_label
TEST Simple validation of freesurfer.read_label behavior
BF Add correct parameters to test
Use HTTPS for outside scripts/forms
DOC http->https where both work and have same content
DOC: Update URLs with non-trivial changes
NF: Automatically reshape FreeSurfer ico7 niftis
TEST: FreeSurfer ico7 surface NIfTI hack
TEST: Update nitest-freesurfer submodule
STY: Remove blank lines from test_freesurfer_ico7_hack
BF: Use deprecate_with_doc as decorator
BF: Fix benchmark range bounds
PEP8/PyFlakes cleanup of nibabel/ (tests pass)
PEP8/PyFlakes cleanup of nibabel/*/ excluding tests (tests pass)
BF: FreeSurfer nifti surfaces can have >3 dimensions
TST: Call current python executable
TEST/BF: Test cases skipped; STY: PEP8
TEST: Update out_tup to what is expected
TEST: Clarify test_qform_sform test
BF: Misplaced parenthesis
TST: Verify pixdim check for non-negative values
RF: Begin refactoring load into image classes
Relegate slicing to is_header, remove _minctest
TST: Remove Minc1Header exception
STY: Unnecessary try block, comment lines too long
TST: Test ValueErrors raised by is_header
RF: Rename is_header to may_contain_header
RF: Restore and deprecate loadsave helpers
TST: Kill extra test; doc string to comment
RF: Prefer alternate_exts over abusing files_types
DOC: Deprecation warnings to stacklevel 2
STY: Remove misleading 'sizeof_hdr' from Minc1/2
RF: Tag sniffs with file name
RF: Fail quietly on too-short header blocks
TEST: path_maybe_image part of image API
BF: Set strided_scalar as not writeable
STY: PEP8 cleanup
valid_exts is no longer used
DOC: Add usage to ImageOpener docstring
STY: PEP8 and import cleanup
BF: Prepare for pydicom 1.0
TST: Travis with pydicom > 0.9.9 git commit
TST: Add sdist, wheel & requirements INSTALL_TYPEs
TST: Add large nifti file test
ENH: Support calling Sphinx with Python 3
DOCTEST: Python 3 fixes
DOCTEST: Use texext 0.3
DOC: Update docs from 2.0.2 release
DOC: Add maintainer, escape $ in RST
DOC: Update website/mailing list
TST: Run style tests separately
RF: Switch to locally-defined underscore function
BF: Update Cifti2Label._to_xml_element
TST: Exercise Cifti2 classes more thoroughly
BF: Add label_table to named_map structure
DOC: Update write_morph_data docstring
TST: Test morph_data round trip
DOC: Update docstring
RF: Enable writing morphometry data to file-like objects
TST: Test dimension, size, and fnum constraints
TST: Restrict accepted shapes to plausible vectors
TST: Skip big array check on 32-bit systems
DOC: Update docstrings
TEST: Actually test oldest matplotlib
DOC: Fieldtrip moved from Google Code to GitHub
NF: Handle metadata in read/write_geometry
BF: Coerce parsed GiftiDataArray.ext_offset to int
TEST: Test default types for parsed Gifti files
BF: Handle empty strings in ExternalFileOffset
RF: Switch to locally-defined underscore function
BF: Update Cifti2Label._to_xml_element
TST: Exercise Cifti2 classes more thoroughly
BF: Add label_table to named_map structure
DOC: Update docstrings
Address comments
TEST: Move shape API tests into DataInterfaceMixin
Update authors
Add new release skeleton to Changelog
Summarize changes since 2.1.0
Update .mailmap
Replace execfile with runpy (2.7 & 3.2+)
PR 504 not merged yet
Import BytesIO from io where possible
Add ArrayProxy.reshape
Typo
Cifti2 arrayproxy tests should be negated or dropped...
RF: Do not automatically load data CIFTI data
Severely simplify _Cifti2AsNiftiImage
Permit tuple ArrayProxy spec in place of header
ArrayProxy.reshape generates headerless ArrayProxy
RF: Deprecate ArrayProxy.header to be removed in 3.0
TEST: Test tuple specs and reshaped APs
Update Changelog
DOC: Update ArrayProxy docstrings
RF: Update ArrayProxy.reshape to take -1 entries
FIX: Correct reshape method
TEST: Catch warning, test -1 reshaping
TEST: Check for illegal specs
TEST: Check legal but incompatible shapes
Update pre-release changelog
Update authors, mailmap
Update numpy minimum version
RF: Derive affine in SpatialImage.transpose
Update docs, function name
DOC: Update changelog, authors
DOC: Update Changelog to include pr/551
DOCTEST: Use legacy float array printing for now
STY: Hide testsetup from docs
TEST: Create and import setup_test fixture into affected files
ENH: Update min numpy version for legacy print
TEST: Import test_setup into module __init__
STY: Ignore unused import
DOC: Update author list
REL: Update version for final tests
DOC: Credit doc fixes
DOC: Update Changelog
TEST: Use label file that exists across FS versions
2.2.0 Changelog
Start 2.2.x maintenance branch
Bump dev version
FIX: Set L/R labels in orthoview correctly
TEST: Check only relevant warnings
STY: Narrow exception check
STY: No bare excepts
Reset maintenance version
MAINT: Update changelog, authors
MAINT: Remove -dev from version
MAINT: Bump dev version for 2.2.x
FIX: Use SSH address to use key-based auth
FIX/ENH: get/set_data_shape more robustly
RF: Subclass MGHHeader from LabeledWrapStruct
RF: Remove one-line private method
FIX: goodRASFlag check, update
RF: Remove unused _load_cache
FIX: Label x/y/z_ras correctly
RF: Make direction cosines explicit, not transposed
RF: Pxyz_c to c_ras
RF: Simplify to/from affine
Rename goodRASFlag -> ras_good
FIX/TEST: MGHImages must be 3D or 4D
RF: delta -> voxelsize; simplify conversion code
TEST: Update tests for refactor
DOC: Annotate header fields
RF: Improve footer handling
ENH: Return TR as 4th zoom for 4D images
RF: Reset zooms to 1 on reshape
TEST: Footer; FIX: Simplify defaults
FIX: Python 2 division
TEST: Missing functionality tests
ENH/DOC: Update and document get/set_zooms
RF: Simplify affine conversion code
ENH: Check for expected byte order
ENH: Re-add _header_data with deprecation warning
FIX: Return on get, not on set
REVERT renamed/reshaped header fields
RF/TEST: Check endianness rather than ignoring
RF: Remove endianness, override copy, as_byteswapped, diagnose_binaryblock
FIX: Do not double-set attributes
TEST: TestWrapStruct features, set_zooms, deprecated fields
TEST: Add and test version check
TEST: Test MGHHeader._HeaderData fully
STY: Typo
DOC: Add reference to FreeSurfer wiki
STY: Whitespace
STY: Remove spaces after final newline
TEST: Test angle_axis2mat(is_normalized=True)
TEST: Check keep_file_open="auto"
TEST: Test deprecation, ornt_transform error condition
TEST: Test voxel size padding
TEST: Check invalid caching argument for get_fdata
FIX: hasattr uses getattr, not __hasattr__
TEST: Test package info functions
TEST: Test from_header(check=False)
ENH: data shape check preempted by update_header
FIX: Give Windows time to decide files are closed
Revert "FIX: Give Windows time to decide files are closed"
TEST: Delete arrayproxies, allowing Windows to close filehandles
COV: Omit tests, drop obsolete omissions
CI: Remove duplicate Travis entry [skip ci]
DOC: Update Changelog
REL: Update version
DOC: Update copyright, contributors
DOC: Update changelog pre-emptively
DOC: Another pre-emptive changelog entry
ENH: Add funcs.crop_image
ENH: crop_image to bounds, adding margin
STY: Do not capture exception in variable
ENH: Permit cropping with negative indices, give more specific errors
TEST: Initial crop_image test
ENH: Enable SpatialImage.{slice,__getitem__}
ENH: Enable minimally-fancy indexing
RF: Move to img.slicer approach
RF: Simplify affine transform
REVERT fileslice modifications
ENH: Add SpatialImage._spatial_dims slicer
RF: Move slicer validation into image
RF: Remove crop_image (for now)
FIX: Type check, _check_slicing arguments
ENH: Handle new axes in Minc
TEST: Image slicing
Translate ValueErrors to IndexErrors when used
TEST: Handle MGHFormat (max: 4D) correctly
TEST: Test single-element list indices
RF: All SpatialImages are makeable
TEST: Check step = 0 case
ENH: Drop fancy indexing
TEST: Test single-slice and fancy indexing raise errors
DOC: Changelog entry
TEST: Do not use deprecated _header_data
DOC: Link to aliasing wiki article [skip ci]
ENH: Improve exception message
RF: Specify spatial dims in each class
TEST: Check for _spatial_dims
TEST: Minor fixes
RF: Switch to spatial_axes_first
RF/TEST: Purge _spatial_dims and related tests
RF: Move slicing machinery to SpatialFirstSlicer class
ENH: Raise IndexError on empty slices
TEST: Update slice_affine test, check zero slices
RF: Fuse only available on systems with getuid
DOC: Some clarifying comments in SpatialFirstSlicer
DOC: Update changelog to include gh-614
DOC: Update changelog
Bump dev version
MAINT: 2.4.0-dev
DOC: Update changelog with release date
DOC: Update changelog with release date
TEST: Ensure warning isn't suppressed
CI: Test on Python 3.7 on Travis
CI: Use dist xenial
CI: Use matrix include for Python 3.7
CI: Use Python 3.7-dev
DOCTEST: FutureWarning repr changed in 3.7
CI: Use 3.7 proper, move trusty+3.4 to matrix
TEST: Catch warnings in clear_and_catch tests
CI: Add Python 3.7 to AppVeyor matrix
FIX: Annotation labels are RGB, no alpha
TEST: Update tests to use RGB only
DOC: Update ctab documentation, comments
CI: Drop extraneous 2.7 build on Travis
DOC: Update Changelog
MNT: .mailmap
MAINT: Drop -dev
MNT: Update YOH, CC initials
TST: Require hypothesis>=3.7
STY: Break overly-long line
ENH: Add ndim to ArrayProxy protocol
TEST: Validate expected ndim behavior
ENH: Pass ndim through to DataobjImage
TEST: Validate DataobjImage.ndim
Revert "TST: Require hypothesis>=3.7"
DOC: Update changelog
Update changelog
REL: 2.3.1
MNT: 2.3.2-dev
Update changelog on master
MNT: Remove deprecate numpy.asscalar
STY: Flake8 3.6+ cleanup
FIX: Immutable buffers should not be set writeable
Revert "FIX: Immutable buffers should not be set writeable"
MAINT: Add bz2file dependency for Python 2
RF: Drop SAFE_STRINGERS, depend on bz2file to provide .readinto()
CI: Upgrade setuptools
MAINT: Move to setuptools exclusively
FIX: Use bz2file for Python 2
FIX: Rely on openers.BZ2File to provide correct class
CI: Upgrade pip, setuptools, wheel on AppVeyor
TEST: fobj string assumptions need only apply to bytearrays
FIX: Load TckFile streamlines using bytearray
TEST: Return mutable bytearray from trk_with_bytes
FIX: Truncate buffer on EOF
CI: Try alternative pip install
MAINT: Bump -dev version
FIX: Use io.BytesIO to test TckFile
STY: Drop unused variable
RF: Leave nisext alone, pop install_requires
DOC: Add a comment explaining readinto/frombuffer idiom
DOC: Update changelog
DOC: Update authors, copyright year
MAINT: Version 2.3.2
Update nibabel/cifti2/cifti2.py
Update nibabel/externals/oset.py
MAINT: Add six dependency
CI: Improve Travis test conditions
MAINT: Bump maintenance version
DOC: Update changelog
REL: 2.3.2
DOC: Changelog typos
MAINT: Bump dev version
RF: Reduce concatenations further, moderate cleanups
FIX: Return to bytearray/frombuffer approach
FIX: Check final delimiter is ONLY infs
TEST: Simple load test for matlab_nan.tck
STY: Reduce diff
RF: Restore missing streamline delimiter error
STY: Pacify flake8
RF: Move setup_test to nibabel/__init__.py
RF: Use pkgutil to avoid importing nose
RF: Check for mock using pkgutil
FIX: Import numpy in setup_test
FIX: Ensure nose can be imported during mock test
TEST: Extending ArraySequences corrupts sliced data
CI: Run coverage on as many tests as possible
RF: Use _safe_resize helper function
CI: Ignore tests, benchmarks and externals for Codecov
CI: Restore coverage config, remove codecov config
CI: Run coverage on Windows
MNT: Drop coveralls - redundant with codecov and poorly configured
DOC: Add Zenodo badge to README
MNT: Add .zenodo.json, tool to sort by commits
MNT: Sort .mailmap
TEST: Move test DICOMs into nicom/tests/data
TEST: Restore dicom files used for DFT
MNT: Last, First name
MNT: Update zenodo metadata, re-sort
RF: Remove unnecessary try guard
DOC: Changelog skeleton
MNT/DOC: Update mailmap, zenodo, author list
MAINT: Version 2.4.0
TEST: Check world direction of encoding remains consistent
RF: Use list comprehension in reordering
ENH: Add NIFTI_XFORM_TEMPLATE_OTHER xform code
MNT: Update author list, mailmap, zenodo
DOC: Update changelog
Apply suggestions from code review
MNT: Bump minimum numpy version to 1.8
TEST: Add de-identified, compressed DICOM with empty string in SliceThickness field
CI: Attempt to use virtualenv on Windows
CI: Use pypiwin32 over pywin32
Apply suggestions from code review
CI: Maybe no update to pywin32 is needed?
Update nibabel/cifti2/cifti2.py
Apply suggestions from code review
DOCTEST: Drop doctest SKIP directives
FIX: Use other.volume_mask to index other.voxel
DOC: Update docstrings with a few more links and array_like
DOC: Update changelog, zenodo
DOC: Add Matt Cieslak to author list
MNT: Bump version to 2.4.1dev
MNT: Bump version to 2.5.0dev
DOC: Update Sphinx config to support recent Sphinx/numpydoc
TEST: Verify writing to loaded data array fails
RF: Decode B64* data to bytearrays for writable numpy arrays
MAINT: Update Zenodo, Changelog with errors/missing info
FIX: Replace invalid MINC1 file with fixed file
TEST: Check for competing actions on warning filters
FIX: Safer warning registry manipulation when checking for overflows
RF: Use catch_warnings context managers to catch overflow
TEST: Clarify comments
MAINT: Pacify upstream Deprecation/FutureWarnings
RF: De-duplicate - the vulnerable window stays small
TEST: Use realistic data in test, loop to ensure interference
MAINT: Deprecate keep_file_open == "auto" (scheduled for 2.4.0)
TEST: Test mmap parameter acceptance
RF: Move mmap/keep_file_open parameters to DataobjImage.from_file*
FIX: Pass kwargs from Cifti2.from_file* to Nifti2.from_file*
FIX: Accept mmap keep_file_open for Ecat and Minc, with no effect
DOC: Update changelog, author list
TEST: Verify we can write files with repeated labels
FIX: Correctly write .annot files with duplicate labels
TEST: Add round-trip test
DOC: Update long description to include CIFTI-2, BRIK/HEAD
DOC: Update changelog
DOC: Update author list
DEP: Deprecate nibabel.trackvis, effect old FutureWarning threats
DEP: Transition checkwarns to deprecations and set target version
DEP: Transition minc module to deprecated, announce 3.0 removal
DEP: Promote catch_warn_reset FutureWarning to DeprecationWarning
TEST: Use old-style assert_warns
TEST: Create a test file for scheduling removals
MAINT: Set name to nibabel in setup.py
MNT: Bump version to 2.4.2dev
TEST: Generalize, schedule nibabel.checkwarns removal
TEST: Verify ImportError, add object removal schedule, schedul trackvis removal
PY2: No importlib
CI: Fix AppVeyor codecov paths
FIX/TEST: Numpy casting rules have gotten more strict, raise different exception
ENH: Add FileBasedImage.serialize()
DOC: FileBasedImage.serialize.__doc__
FIX: Import io
ENH: Test serialize via an API test mixin
RF: Rename serialize to to_bytes, add from_bytes
RF: Factor SerializableImage
TEST: Generate file when no examples available
TEST: klass sometimes missing, equality sometimes undefined
ENH: Add to/from_bytes interface to GiftiImage
TEST: Delete images so mmapped files can be removed
DOC: Improve docstring
FIX: Nifti1Image is serializable, not Nifti1Pair
ENH: Check for multi-file images, to ensure well-defined behavior
TEST: Move special case header equality check into TestSerializableImageAPI
STY/DOC: Note about multi-file images, newline
MAINT: Full-speed setuptools
MAINT: Restore reading version from info.py
PY2: ConfigParser for Sphinx build
WIN: Drop .bat suffix for scripts
MAINT: externals/tests/data skipped in sdist
MAINT: Add minimum setuptools version to setup.py
CI: Test minimum setuptools
MAINT: Add pyproject.toml to give setuptools hint
MAINT: Add wheel requirement on-the-fly, improve comments
TEST: Add multiframe ECAT file for testing
FIX: Minor ECAT cleanups
DOC: Update changelog for upcoming 2.5.0 release
MAINT: Version 2.5.0
FIX: Set TRK byte order to < before filling values
PY3: Use next function on iterator while checking ECAT affines
DOCTEST: Display system endianness-neutral representation of <i2
TEST: Use package-wide setup and teardown to adjust numpy print options
MAINT: Use a more general _test_state structure for setup/teardown
Apply suggestions from code review
ENH: Pass endianness as parameter to TrkFile._default_structarr
DOC: Update changelog
MAINT: Update mailmap, Zenodo
REL: 2.5.0
MAINT: 2.5.1-dev
MAINT: 3.0.0-dev
MAINT: Add community guidelines
DOC: Update docstring standard
DOC: Contributing guide, mostly linking to docs
DOC: Drop covenant boilerplate
DOC: Add Matthew Brett as point of contact
CI: Add Azure Pipeline for Python 3.5
CI: Remove AppVeyor
AZURE: Environment reset at each step
AZURE: Virtualenvs do not seem to work as expected
AZURE: Fix step name
AZURE: Expand matrix to Python 3.4-3.7, x86 and x64
CI: Test installing from git archive
CI: Drop 2.7, 3.4 tests from Travis
MAINT: Update dependencies for Python >= 3.5
MAINT: Add update_requirements.py; update requirements
FIX: Update minimum six
FIX: Need numpy 1.12 for fixed assert_warns
MAINT: Purge __future__
RF: Drop six.*metaclass
RF: Drop six.*IO
RF: Drop six.moves
RF: Drop six.*_types
MAINT: Drop nibabel.externals.six
MAINT: Drop six dependency
RF: Drop code explicitly tagged as PY2
RF: Drop python < 3.5.1 hacks in openers
TEST: Duplicate assertion
DOC: Update docstrings without Py2/3 discussion
RF: Remove Python < 3.5 hacks
TEST: Drop BZ2 mio error checks (Python 3.3 only)
RF: Use builtin FileNotFoundError
STY: Unused import
RF: Import from numpy.compat.py3k when needed
TEST: Minimum numpy >= 1.12
DOC: Minimum Python == 3.5.1
MAINT: Set up versioneer
MAINT: Drop COMMIT_INFO
RF: Drop old __version__
RF: Move all version play from info to pkg_info
MAINT: Add fallback version to versioneer + git-archive
TEST: Test fallback version
VERSIONEER: Improve short version file handling
MAINT: Ignore _version.py in coverage
MNT: Deprecate DataobjImage.get_data, schedule for removal in 5.0
CI: Drop separate indexed_gzip test
STY: Configure pep8speaks
DATA: Add nibabel-data/nitest-dicom submodule
TEST: Update test DICOM file location
MNT: Update extras
CI: Refactor Travis to separate steps more cleanly, use extras to for dependencies
MNT: setup_egg.py now unused
Trigger CI
CI: Add expected failing test
CI: Temporarily drop all other tests
ENH: Require h5py 2.10 for Windows + Python < 3.6
TEST: Import have_h5py and similar from nibabel.minc2
FIX: More robust floating point dtype check
Revert "CI: Temporarily drop all other tests"
CI: Require binaries for 3.4 as dependencies stop building wheels
RF/TEST: Move h5py compatibility hacks into a private module
DOC: Update docstrings to describe affine property
CI: Add expected failing test to Azure
ENH: Add general test data retriever
DOCTEST: Retreive a surface file using test_data
DATA: Add 10 time point time series GIFTI in fsaverage3 space
TEST: Test new test_data function
TEST: Check for failing GIFTI encodings
FIX: Coerce data types on writing GIFTI DataArrays
MAINT: Check nightly builds on 3.7
MAINT: Check nightly builds on 3.7
TEST: Reproduce gh-802
ENH: Remove img.get_data() from internal use except for appropriate tests
DOC: Update test_image_api docstring for clarity, consistency
DOC: Update changelog for upcoming 2.5.1 release
MAINT: Version 2.5.1
MAINT: Update zenodo.json
DOC: Add Henry Braun to contributor list
MAINT: Update setup.cfg nose version to match installation instructions
MAINT: 2.5.2-dev
Re-import externals/netcdf.py from scipy
DOCTEST: Avoid fixed ID in doctest
DOCTEST: Delete reference to mmap data to avoid warning
Re-import externals/netcdf.py from scipy
DOCTEST: Avoid fixed ID in doctest
DOCTEST: Delete reference to mmap data to avoid warning
CI: Test builds on Python 3.8
CI: Python 3.8 not yet available on Azure
CI: Wait on upstream releases for full 3.8 tests
MNT: Update supported language classifiers
DOC: More comprehensive agg_data examples
CI: Restore Python 3.8 Azure tests
NF: Enable data scaling within the target dtype
RF: Implement AFNIArrayProxy._get_scaled(), replacing other overrides
ENH: Add get_scaled to Minc, Ecat
RF: Rework PARRECArrayProxy
TEST: Revise equality check to approximate
CI: Enable most dependencies for Python 3.8
CI: Test with matplotlib~=3.2.0rc1
TEST: Validate get_scaled with several float types
FIX: Better type promotion
TEST: Use np.sctypes to detect available floats
RF: Uniformize AFNI/PARREC scaling
RF: Use consistent logic for ECAT, Minc1 and PARREC proxies
TEST: Verify get_scaled works sensibly with ints
MNT: Update nitest-dicom pin
TEST: Basic tests for CT DICOM image
RF: Avoid unneeded reference
DOC: Update docstrings, comments, for clarity
MNT: Update mailmap
MNT: Update Zenodo
DOC: Update changelog
MNT: Improve version determination on branches with digits
DOC: Update author list
MNT: Version 3.0.0rc1
MNT: Version 3.0.0rc1.post.dev
MNT: Add gitpython and twine to "dev" extra
MNT: Fix Dorota Jarecka ORCID
MNT: Drop test data from MANIFEST.in
DOC: Discuss slicer interface
CI: Update release wheel repository
CI: Re-add h5py to Py38 tests
CI: Restore old wheel repository
TEST: Check non-integral slopes, intercepts
CI: Disable pytest for now
TEST: Test supported dtypes for Analyze/NIfTI/MGH
TEST: Refine parameter and assertion precision
FIX: Cast scaling parameters only when safe
TEST: Add comment to explain parameters
FIX: Accept dtype parameter to ArrayProxy.__array__
TEST: Validate dataobj.__array__(dtype)
RF: Drop (mostly) redundant ArrayProxy.get_scaled() method
DOC: Update changelog
TEST: Filter complex warnings
TEST: Improve test naming for tracking down failures
FIX: ECAT data must be coerced after reading
DOC: Update reviewers in changelog
DOC: Update changelog with recent PRs
MNT: Remove support for keep_file_open=="auto"
MNT: Remove minc.py
DOC: Use nifti1 to demonstrate ModuleProxy instead of the ephemeral minc
MNT: Remove checkwarns
TEST: Test invalid KEEP_FILE_OPEN_DEFAULT values
STY: Remove unused import [skip ci]
TEST: Move from mock to unittest.mock
MNT/DOC: Add h5py to notes, installation extras
DOC: Update changelog
REL: 3.0.0rc2
MNT: Version 3.0.0rc2.post.dev
REL: 3.0
MNT: Begin development on 3.1
MNT: 3.0.x maintenance branch
MNT: 3.0.x maintenance branch
FIX: Validate ExpiredDeprecationErrors
FIX: Interpret "ver+extra" as post-release, not pre-release
TEST: Numpy changed longdouble str representations in 1.18
FIX: Expand version comparison logic
TEST: Numpy changed longdouble str representations in 1.18
RF: Improve fallback version check, require PyPA packaging module
MNT: All supported numpy have float16
MNT: Update changelog, zenodo, in case of future 2.5.x release
MNT: Add Ben Darwin to Zenodo
DOC: Update changelog, copyright
MNT: Update requirements
DOC: Update author list
DOC: Add missing entry from 3.0.0 changelog
MNT: Update Ben Darwin name/affiliation
MNT: 3.0.2-dev
NEP29: Bump minimum numpy to 1.13
MNT: Minimum packaging of 14.3
MNT: Separate out pytest and nose jobs
TEST: Fix tests broken in the rebase
CI: Add CHECK_TYPE to Azure
CI: Default to nose for Windows for now
RF: Avoid running data script code on doctest
TEST: Doctests not ready
CI: Try alternative CHECK_TYPE test
PIN: Avoid pytest 5.3.4 (pytest-dev/pytest#6517)
TEST: Add BaseTestCase class to skip TestCases starting with ``_``
TMP: Skip failing streamlines modules
CI: Skip converted tests in nose
TEST: Fix import/indentation errors
ENH: Use generic unittest.SkipTest in missing optional_package
TEST: Cleanup test_array_sequence
CI: Skip more nose tests
TEST: Un-skip tractogram tests, style cleanups
TEST: Fix optpkg test
TEST: Style updates to test_analyze
CI: Give up trying to deselect streamlines
TEST: Style cleanup
TEST: Simplify data/environment fixtures
TEST: Simplify, reformat test_data
STY: Remove unused pkgutil import
TEST: Refactor NetCDF tests to be more pytest friendly
CI: Resolve Travis warnings, remove redundant job
TEST: Use skipif mark instead of TestMinc2API.__init__
TEST: Use standard syntax when possible
TEST: Use relative imports in gifti tests
TEST: Simplify giftiio deprecation tests
TEST: Combine imports
MNT: Mark incompatible with Python 3.9
TEST: Skip giftiio tests in nose
MNT: CI YAML formatting, extraneous .coveragerc
TEST: Style and formatting
CI: Restore import-only test
CI: Enable doctests
TEST: Use numpy 1.13 legacy print settings
TEST: Convert nibabel.nicom to unittest/pytest
TEST: Use vanilla unittest.skip* over pytest
RF: Remove unused pieces of testing_pytest and switch to unittest skip
CI: Only ignore files that use fixtures
TEST: De-indent tests, cleanup style
TEST/STY: Cleaner line breaks and alignment for test_fileslice
TEST/STY: Alignment in fileutils, floating tests
TEST/RF: Reduce redundancy in test_removalschedule, test failure case
MNT/FIX: Bump minimum matplotlib to 1.5.2 to match available wheels
MNT: Use version 1.5.3 to smooth transition to 3.6
TEST: Remove last imports of nose
MNT: Purge nose from requirements, CI
RF: Remove nose test state code
RF: Reimplement nibabel.test() and nibabel.bench()
DOC: Replace nose with pytest in the docs
DOC: Update benchmark docstrings
TEST: Final few assert_raises
CI: Add OSX test to Travis
CI: Fix wheel build... weird pip behavior while in source directory
FIX: Return pytest.main results from nibabel.test/bench
TEST: Might as well test the testers. Found a bug.
CI: Install test requirements through extra dependencies
DOC: Specify current benchmark file
TEST: Style and parameterization fixes for readability
TEST: Bad test syntax
TEST/RF: Mock compress_ext_map, instead of fixture
TEST: Rename tests to ensure running
TEST/RF: Reparametrize tests
CI: Fail test if docs do not build
PY3: Drop Python 2 guards, use runpy instead of exec
DOC: Attempt to find versioneer version when building docs
DOC: Skip py3k
DOC: Fix "``s" pattern in changelog
DOC: Address warnings and improve formatting
MNT: Update changelog, author list and Zenodo
MNT: Drop pyproject.toml for now
DOC: Note pyproject.toml removal in changelog
MNT: 3.0.3dev
NEP29+1y: Drop Python 3.5
CI: Drop 3.5 tests, set min to 3.6, pre-release to 3.8
MNT: Drop deprecated ndarray.tostring for tobytes
TEST: Fix warning filter to only check for nibabel-generated warnings
MNT: Pin sphinx <3
MNT: Mark incompatible with numpy 1.19, Sphinx 3
MNT: Set maximum numpy to 1.16.x, update requirements.txt
CI: Avoid installing too new of numpy/scipy
REL: 2.5.2
MNT: 2.5.3-dev
NF: Add nibabel.affines.rescale_affine function
RF: Update conformation to reorient, rescale and resample
MNT: Update mailmap
DOC: Update minimum Python
DOC: Update author list
DOC: Update changelog
MNT: Update Zenodo snippet
MNT: 3.1.1-dev
MNT: 3.2.0-dev
MNT: Add affiliation/ORCID for Roberto
CI: Require pip <19.2 for Python 3.4 in Travis
STY: Satisfy recent flake8
CI: Fix Azure pipelines config
CI: Add ARM64 build
CI: Drop ARM64 build dependencies to numpy for speed
CI: Use scipy-wheels-nightly index on anaconda.org
FIX: Ensure nan is cast to currently tested dtype
Update nibabel/orientations.py
Update nibabel/tests/test_quaternions.py
MAINT: Set minimum numpy to 1.14
TEST: Drop numpy 1.13 doctest compatibility
TEST: Update doctests to use new numpy print format
REL: 3.1.1
MNT: 3.1.2dev
RF: Set NIfTI-2 header shape through API call
DOC: Add the spec text to explain the dim discrepancy
FIX: Save TR in seconds in parrec2nii
TEST: Verify that header conversion preserves units
ENH: Add bytestring serialization to CIFTI-2
TEST: Add SerializeMixin to TestCifti2ImageAPI
MNT: Officially deprecate onetime.setattr_on_read
DOC: Update link to descriptor guide
TEST: Suppress expected warnings
FIX: Avoid/suppress expected warnings
FIX: Bad deprecate_with_version calls
PY36: Backport nullcontext
DOCTEST: Suppress deprecation warnings in doctests of deprecated methods
TEST: Suppress warnings in freesurfer
MNT: Add pytest-doctestplus to ignore warnings in doctests
TEST: Suppress and check warnings in streamlines
TEST: Use get_fdata, check for expected warning
RF: Drop calls to deprecated get_affine()
FIX: Escape backslashes before parsing strings
MNT: Reset deprecation clock
FIX: Remove doubled-up fix for divide-by-zero warning
TEST: Small cleanups
FIX: Construct sliceobj from object array; addresses numpy deprecation warning
TEST: Use skipIf rather than warn + return for test_multiload
TEST: Check import-time warnings on modules via conftest
TEST: Use matches or comments to make clear expected warnings
TEST: Unused imports
TEST: Simplify lazy_load conditional warning check
Update nibabel/streamlines/tests/test_tck.py
ENH: Add nib-roi command to crop (maybe flip) axes
ENH: Cover edge cases like non-scaling images, bad slices
TEST: Test nib-roi and helpers
FIX: Syntax errors/warnings
Update nibabel/cmdline/tests/test_roi.py
TEST: Check MINC-2, Analyze, and main()
TEST: Use MGHImage to avoid scipy dependency
FIX: Improved test for scaling
MNT: Update mailmap
DOC: Note new minimum Numpy
MNT: Update author list, zenodo ordering
DOC: Changelog
MNT: 3.2.1-dev
Trigger builds on maint/ branches and PRs
DOC: Update Julian Klug's Zenodo fragment
CI: Test on Python 3.9 with numpy nightly wheels
PY39: Changed mmap repr
PY39: Changes in ast parsing of indices
CI: Bump ARM64 tests with compatible packages
CI: Drop defunct rackspace wheelhouse
CI: More dependencies under 3.9
DOCTEST: Fix np.float uses in the doctests
DOC: 3.2.1 changelog
MNT: List Python 3.9 in supported versions
FIX: Drop keep_file_open=auto from benchmark
Christopher P. Cheng (95):
TODO1 attempt 1: processed data type, data shape, and data offset of two different headers
tweaked to remove AnalyzeHeader but currently still has problems
added nib-diff to setup.py
first attempt at nib-diff that doesnt work
removed incorrectly committed changes
first attempt at diff_dicts method and diff testing file
tried something else with header fields
latest attempt: restructured diff_dicts() method, troubleshooting ensuing errors
corrected misplacement of cmdline files and latest attempt at diff_dicts method
progress! tweaked bugs, corrected rookie mistakes like cmdline placement, implemented yarik suggestions
got rid of proc file and function works at a basic level
tweaked diff_dicts to be compatible with tests
got rid of None, troubleshot tests
introduced hypothesis to use for testing with pretty sexy results
noted hypothesis need for tests, refactored diff_dicts name
attempt at TODO#2: allowing specification of header fields
now functional for several header files.
tweak to make hypothesis work with a list, but problem above has not been addressed
tweaked names and code as suggested!
bug fix
cosmetic tweaks
cleaned up code
promoted generic programming and got test to work again
tried to clean code but couldnt get comprehension going
comment and docstring
added options for text, json, yaml but still have to implement
work in progress with all outputs
moved nib-ls to make generic use of its functions possible for a comparison function in progress
fixed import from utils to ls
started on a test_utils module. Not the final tests, just some random strings
changes as recommended, thanks Matthew!
added test for ap() and made changes re Travis CI
fixed ap test problem but TypeError from nose popped up
format is out of line but problem is fixed
simplified code but now bizarrely doesnt run with two files
apparently the variable l is a really big chafe
beautified text output, next up json and yaml
modified the tests for the new diff_values
commenting out json and yaml, cosmetic mod to ls.py
made changes as requested
moving more cmdline things to the cmdline folder
progress towards table formatting
table now works and looks p good
slight tweak to table formatting
very inefficient but successfully removed dtype when its identical
removed extra blank line lol
boosting coverage
refactoring
added dicomfs file
parrec2nii added
data comparison function added
reformatted per Travis CI
implemented test-nib-diff
added another test
changed travis appveyor utils per test errors
test files
wasnt included in last commit by accident
maybe this will work instead
maybe this will work instead
test fixed
fixed to be more generic
fixed file name issue in nib diff
fixed changes as yarik requested
fixed problems noted by Travis CI
building in difflib to see whats going on
latest attempt at compliance changes dict to list
combined into one line
new structure, names, tests
hypothesis: all my problems were due to that one test
whoops missed this
i think im going to cry. code cleaned and made more pythonic
added and fixed tests
fixing up appveyor and travis problems
fixed a fringe use case
style tweak for travis
removed duplication, made things more generic
moved functionality outside to test and increase coverage
boosting coverage by testing main
main test corrected for max coverage
imported StringIO from six instead of io
added a test for cmdline function
changes per Yariks comments!
changed as commented out in the pull request
RF: anticipated files of different shapes, fixed table display, corrected tests
elaborated docstring, modified get_data_diff to allow direct array input, added tests for coverage
added to diff documentation, undid executable change, took out debugging script in test_scripts
undid permission snafu on test_scripts
docstring and function name clarification, change get_data to get_fdata()
corrected styles per Travis, limited fdata to float32
prepared for future PR to allow modification of dtype used in diff comparison
added cmdline functionality for modifying datatype used for file data comparisons
added testing for dtype at cmdline, assuming cmdline functionality. changed value output in data_diff function to allow for validation of data type
storing changes for main testing
added tests to fix coverage problem
cmdline test added for dtype access
Cindee Madison (27):
NF - initial import of ecat functionality
NF added correct frame mapping to mlist class
update EcatImage to work with new nibabel changes
NF clean up for Ecat class
NF warning if affines do not match across frames
NF further cleanup of code, references to mlist cleaned up
NF fix warning in fileobj seek
NF adding ecat tests, fix endian typo
NF more tests fix issue in endian in tinypet.v
NF more tests fix issue in endian in tinypet.v
NF tests for EcatSubHeader
NF tests for EcatImage
NF: added full datatypes, filetypes descriptions , documentation
NF: endian issue with tinypet.v
NF: added access to mlist and subheader from img, and exceptions for saving
NF: added simple examples
BF - fixed doctests in ecat
BF fix doctest for ecat
RF - removed parametric tests from test_ecat
BF - fix endian test fail on BE machines
fix bug in identifying correct frame number order in a multi-frame ecat file using mlist
addd test to find misorder in mlist
added test for poorly formed mlist order
fix typo in get_frame_affine
clarified mlist.get_framenumbers based on comments from MB, added new mlist function get_series_framenumbers to give series frame numbers for data given in file
added examples to new function get_series_framenumbers
BF+TST catch mismatched hdr, mlist and added more tests
Demian Wassermann (58):
BF: Fix file trackvis file read when there are scalars and the point mode is rasmm
TST: test scalars work with rasmm trackvis option
Fix Cifti2TransformationMatrixVoxelIndicesIJKtoXYZ to_xml method
Changed the cifti2.py assertions for exceptions
Changed the parse_cifti2_fast.py assertions for exceptions
Fixed MetaData and its testing
Exporting the CIFTIHeaderException to the cifti2 namespace
Improved verification/compatibility of the label object with the standard. Increased testing coverage
All elements below MatrixIndicesMap fixed and tested
Newline homogeneization
Fixed header information
Cleaned the metaclass helpers
Fixed docstring
Fixed style and function of MetaData
MetaData now behaves like a mapping
Interface of LabelTable closer to that of Sequence
Better agreement between the Label element and its specification
Several docstring fixes as well as removed the squeeze
Code cleaned
Code cleanup
Fixed docstring
Removed comments
Added submodule for CIFTI2 testing
Added test data as submodule
Added more I/O tests fixed load/save functions:
CIFTI2: MetaData and LabelTable now behave like MutableMapping
CIFTI2: MatrixIndicesMap has a MutableSequence behavior
Simplified testing by assuming that the parsed header is correct
CIFTI2 Improved MatrixIndecesMap interface
CIFTI2 Cleanup
CIFTI2 simplifications in implementation and testing due to the new interfaces
Compatibility with new/future nifti interface
Fix testing dependency on lxml
VertexIndices and VoxelIndicesIJK now have a MutableSequence interface
Improved testing coverage
Better handling of volumes in the MatrixIndicesMap
Deleted intent_code restriction
Removed lxml dependency
Increased testing coverage and removed residual methods
Clarified code in _Cifti2AsNiftiImage __init__
Changed Cifti2Vertices such that it behaves like a sequence. Added corresponding tests
Icreased testing coverage. Cleaned up code
Corrected problem in new nifti header checks
Increased testing coverage and code cleanup
Fixed python 2.7 incompatibility in tests
Removed commented code
Remove affine attribute from CIFTI2 image __init__
CIFTI2 reading now is performed without loaading all file from disk
MetaData can be built from a mapping
Improved docstring of Cifti2BrainModel
Improved docstring of Cifti2MatrixIndicesMap
Code clean
Code clean
Renamed parse_cifti2_fast to parse_cifti2
Renamed CIFTI2HeaderError to Cifti2HeaderError
Unified the CIFTI2 spelling in the documentation
Fixed documentation typo
Added convenience functions
Dimitri Papadopoulos Orfanos (1):
Better handle pathological filenames
Dorota Jarecka (104):
moving test_affines to pytest
adding pytest to ravis
adding pytest to the windows env
testing windows installations
trying a different pytest command for windows
yet another command for windows
fixing path to the test
revert some of the changes to the test (didnt notice numpy.testing was already used)
adding testing_pytest that removed nose (it will be used temp. for tests converted to pytest)
converting tests with tests/test_a*.py and test_wrapstruct (for some reason cant run the tests in one pytest command, so added an extra line to travis for now
converting test_b* and test_c*
updating travis
adding converted test_d*
converting more tests to pytest
fixing interaction between fixtures
moving test_helpers to testing/helpers.py (it contains no tests, but helping functions)
adding new tests to travis
adding more tests
converting more tests
converting more tests to pytest
converting test to pytest, TODO: some tests are failing on osx
adding pytestmark to test that has not been changed to pytest
fixing test_wrapstruct: removing setUpClass and adding self.header_class check to all of the test methods: couldnt find better way for suppressing tests from abstract classes and allowing for testing from all child classes
changing travis command after fixes in test_wrapstruct
converting more tests from nibabel.tests
converting nibabel/tests/test_scripts.py
adding todo
converting more tests nibabel.tests.test_s*
converting more tests nibabel.tests.test_t*
converting test_volumeutils
converting nibabel_data
removing todos
Apply suggestions from code review
Apply suggestions from code review
converting nibabel/tests/test_scripts.py
adding todo
removing todos
small edits: suggestions from reviews
ignoring tests with pytest.mark.parametrize
converting nibabel/tests/test_scripts.py
adding todo
converting more tests nibabel.tests.test_s*
removing todos
Apply suggestions from code review
converting nibabel/tests/test_scripts.py
adding todo
removing todos
small edits: suggestions from reviews
applying suggestions from 865
converting nibabel/tests/test_scripts.py
adding todo
converting more tests nibabel.tests.test_s*
removing todos
converting nibabel/tests/test_scripts.py
adding todo
removing todos
small edits: suggestions from reviews
converting nibabel/tests/test_scripts.py
adding todo
converting more tests nibabel.tests.test_s*
removing todos
converting nibabel/tests/test_scripts.py
adding todo
removing todos
small edits: suggestions from reviews
converting nibabel/tests/test_scripts.py
adding todo
converting more tests nibabel.tests.test_s*
converting nibabel_data
removing todos
converting nibabel/tests/test_scripts.py
adding todo
removing todos
small edits: suggestions from reviews
converting nibabel/tests/test_scripts.py
adding todo
converting more tests nibabel.tests.test_s*
removing todos
converting nibabel/tests/test_scripts.py
adding todo
removing todos
small edits: suggestions from reviews
converting nibabel/tests/test_scripts.py
adding todo
converting more tests nibabel.tests.test_s*
removing todos
converting nibabel/tests/test_scripts.py
adding todo
removing todos
small edits: suggestions from reviews
converting nibabel/tests/test_scripts.py
adding todo
converting more tests nibabel.tests.test_s*
removing todos
converting nibabel/tests/test_scripts.py
adding todo
removing todos
small edits: suggestions from reviews
ignoring tests that use needs_nibabel_data
some cleaning after using nose2pytest
small cleaning after review
small edit
Update nibabel/tests/test_image_api.py
changing testing_pytest back to testing
Egor Panfilov (3):
Fixed typo in coordinate_systems doc page
Fixed typo in coordinate_systems doc page
Fixed A-P to P-A in coord systems docs
Eleftherios Garyfallidis (3):
NF: Tractogram is now more specific and supports multiple keys for scalars and properties which are now called data_per_steamlines and data_per_points respectively
Updated TractogramFile
minor cleanup
Eric Larson (64):
ENH: Adding vox2ras_tkr
FIX(?): Add zooms test
FIX: Deal with borked test dataset
FIX: Match MRIRead.m
FIX?: Order
FIX: Reverting to mri_info test compat
FIX: Cleaner tests
FIX: Minor fixes
ENH: Add suppress_warnings
FIX: Support XYTZ order and multiple echos in PAR/REC files
ENH: Support for partial recordings, and overwrite checking
WIP: Add bvals/bvecs
FIX: Working for variable scale factors
ENH: Add dwell time
FIX: Verbose dwell output
FIX: Fix dwell time
FIX: Fix numpy
FIX: Move most work to parrec
FIX: Fix for truncated data
FIX: Fix tests
FIX: Fix openers
FIX: Minor fixes to clean up tests in windows
FIX: Py3k on Linux fixes
STY: PEP8
FIX: Remove trace by default
ENH: Add appveyor and improve test
ENH: Add an extra windows build
FIX: Fix badge
ENH: Add crosshairs, modify mode
FIX: Minor fixes
ENH: Add set_indices method
ENH: Allow time dimension
ENH: Better tests
FIX: Update volume plot
FIX: Remove monkey patching
WIP
WIP: Closer to correcting orientation
WIP: Fixed ratio
FIX: FIX time plot
FIX: Fix orientations and interactions
FIX: Minor fixes
FIX: Fix test
FIX: Better testing
STY: Remove dicts in favor of lists
FIX: Cleanup on close
FIX: Better unlinking
FIX: Minor fix
FIX: Fix operators
FIX: Clean up code
FIX: Fix for NumPy deprecation
STY: Clearer conditional
FIX: Allow tr=0 in mghformat
FIX: Better messages
BUG: Fix deprecated imports
FIX: Escapes and NumPy import
BUG: larsoner
BUG: Fix escape
BUG: More escapes
BUG: Fix escape
BUG: More escapes
BUG: Fix escape character
FIX: r
BUG: Use manager to set title
FIX: NumPy builtin type deprecation
Eric M. Baker (7):
BF: Remove assumption that StackID is always the first dimension of the frame indices for multiframe DICOM
BF: Correct StackID DICOM tags
TST: Update test for MultiframeWrapper.image_shape
TST: Update dependent tests that were missed previously
DOC: Add short explanation and references to image_shape function for the Multi-frame DICOM wrapper
RF: Remove use of 'FunctionalGroupPointer' in determining 'StackID' axis
DOC: Fix external link rendering
Erik Kastman (30):
Add Nifti1DicomExtension + test
Add @dicom_test decorator to Pass w/o pydicom
Use BytesIO (py3k compatibility)
Fix BytesIO import (again), py3.3 compat
Fix writing typo, add writing tests
Zeropad Extension to 16 bytes, test writing
A little cleanup
Read full datasets (with TransferSyntax)
Use write_dataset for pydicom < 0.9.9 compat
PEP8 Whitespace
Test TransferSyntax, fix zipped reading.
Revert "Read Full Datasets Using TransferSyntax"
Revert "PEP8 Whitespace"
Revert "Use write_dataset for pydicom < 0.9.9 compat"
Revert "Read full datasets (with TransferSyntax)"
PEP8 + consistency cleanup
Create empty dataset or use one if provided
Add Doc page on NIfTI header
PEP8 and header warning (doc)
Import and doc cleanup
PEP8 Cleanup
NiftiHeader determines dicom byte encoding in extension
DOC Add docstring to DicomExtension __init__
RF Type cleanup from MB’s suggestions
BF Import pydicom…- Loading branch information
