[BUG] annot_to_gifti stopped working

[ltuominen]

Issue summary

Hi there,
I am rerunning my code and since last update annot_to_gifti stopped working. Something about unable to read giftis even if I give it annot's. Appreciate your help! Thanks again
Lauri

Detailed issue description

base_path = os.path.dirname(os.getcwd())
dk_fsaverage_164k = (base_path + '/parcellations/atlas-desikankilliany_space-fsaverage_den-164k_hemi-L.aparc-1.annot',
                     base_path + '/parcellations/atlas-desikankilliany_space-fsaverage_den-164k_hemi-R.aparc-1.annot')
dk_fsaverage_164k = annot_to_gifti(dk_fsaverage_164k)

---

TypeError Traceback (most recent call last)
Cell In[4], line 4
1 base_path = os.path.dirname(os.getcwd())
2 dk_fsaverage_164k = (base_path + '/parcellations/atlas-desikankilliany_space-fsaverage_den-164k_hemi-L.aparc-1.annot',
3 base_path + '/parcellations/atlas-desikankilliany_space-fsaverage_den-164k_hemi-R.aparc-1.annot')
----> 4 dk_fsaverage_164k = annot_to_gifti(dk_fsaverage_164k)

File ~/anaconda3/envs/neuromaps_py3.8/lib/python3.8/site-packages/neuromaps/images.py:529, in annot_to_gifti(parcellation, background)
525 labeltable.labels.append(glabel)
527 gifti += (nib.GiftiImage(darrays=[darr], labeltable=labeltable),)
--> 529 return relabel_gifti(gifti, background=background)

File ~/anaconda3/envs/neuromaps_py3.8/lib/python3.8/site-packages/neuromaps/images.py:456, in relabel_gifti(parcellation, background, offset)
452 background = PARCIGNORE.copy()
454 for hemi in parcellation:
455 # get necessary info from file
--> 456 img = load_gifti(hemi)
457 data = img.agg_data().copy()
458 labels = img.labeltable.labels

File ~/anaconda3/envs/neuromaps_py3.8/lib/python3.8/site-packages/neuromaps/images.py:170, in load_gifti(img)
166 # it's not a pre-loaded GiftiImage so error out
167 elif (isinstance(err, TypeError)
168 and not str(err) == 'stat: path should be string, bytes, os.'
169 'PathLike or integer, not GiftiImage'):
--> 170 raise err
172 return img

File ~/anaconda3/envs/neuromaps_py3.8/lib/python3.8/site-packages/neuromaps/images.py:160, in load_gifti(img)
143 """
144 Load gifti file img.
145
(...)
157 Loaded GIFTI images
158 """
159 try:
--> 160 img = nib.load(img)
161 except (ImageFileError, TypeError) as err:
162 # it's gzipped, so read the gzip and pipe it in
163 if isinstance(err, ImageFileError) and str(err).endswith('.gii.gz"'):

File ~/anaconda3/envs/neuromaps_py3.8/lib/python3.8/site-packages/nibabel/loadsave.py:96, in load(filename, **kwargs)
81 def load(filename: FileSpec, **kwargs) -> FileBasedImage:
82 r"""Load file given filename, guessing at file type
83
84 Parameters
(...)
94 Image of guessed type
95 """
---> 96 filename = _stringify_path(filename)
98 # Check file exists and is not empty
99 try:

File ~/anaconda3/envs/neuromaps_py3.8/lib/python3.8/site-packages/nibabel/filename_parser.py:41, in _stringify_path(filepath_or_buffer)
25 def _stringify_path(filepath_or_buffer: FileSpec) -> str:
26 """Attempt to convert a path-like object to a string.
27
28 Parameters
(...)
39 https://github.com/pandas-dev/pandas/blob/325dd68/pandas/io/common.py#L131-L160
40 """
---> 41 return pathlib.Path(filepath_or_buffer).expanduser().as_posix()

File ~/anaconda3/envs/neuromaps_py3.8/lib/python3.8/pathlib.py:1042, in Path.new(cls, *args, **kwargs)
1040 if cls is Path:
1041 cls = WindowsPath if os.name == 'nt' else PosixPath
-> 1042 self = cls._from_parts(args, init=False)
1043 if not self._flavour.is_supported:
1044 raise NotImplementedError("cannot instantiate %r on your system"
1045 % (cls.name,))

File ~/anaconda3/envs/neuromaps_py3.8/lib/python3.8/pathlib.py:683, in PurePath._from_parts(cls, args, init)
678 @classmethod
679 def _from_parts(cls, args, init=True):
680 # We need to call _parse_args on the instance, so as to get the
681 # right flavour.
682 self = object.new(cls)
--> 683 drv, root, parts = self._parse_args(args)
684 self._drv = drv
685 self._root = root

File ~/anaconda3/envs/neuromaps_py3.8/lib/python3.8/pathlib.py:667, in PurePath._parse_args(cls, args)
665 parts += a._parts
666 else:
--> 667 a = os.fspath(a)
668 if isinstance(a, str):
669 # Force-cast str subclasses to str (issue #21127)
670 parts.append(str(a))

TypeError: expected str, bytes or os.PathLike object, not GiftiImage

Steps to reproduce issue

>>>dk_fsaverage_164k = (base_path + '/parcellations/atlas-desikankilliany_space-fsaverage_den-164k_hemi-L.aparc-1.annot', base_path + '/parcellations/atlas-desikankilliany_space-fsaverage_den-164k_hemi-R.aparc-1.annot')

>>>dk_fsaverage_164k = annot_to_gifti(dk_fsaverage_164k)

Software version

0.0.4+41.g899aae1

What operating system were you using when you encountered this issue?

• macOS
• Windows
• Linux

Code of Conduct

• I agree to follow the neuromaps Code of Conduct

[VinceBaz]

Hi @ltuominen. This appears to be caused by a recent PR in nibabel. We're currently fixing this. We even have the PR ready. I was actually working on this at this exact moment. We'll merge it in a couple of minutes max. I'll let you know when the PR is merged 😄

[ltuominen]

Wow! This is the speed I like to see things happening.

[VinceBaz]

The PR is merged @ltuominen. Let us know if the bug is fixed with the new version 😄

[ltuominen]

Works perfectly! Thanks again

[VinceBaz]

Awesome!

[Turius]

I am having the same issue, even if I am using the latest versions of neuromaps (0.0.4) and nibabel (5.2.0). Would you have any tips on how to solve it? @VinceBaz

[liuzhenqi77]

Hi @Turius, we should have fixed the issue on the main branch, but we have not released an official new version on pypi yet. Could you please update the package to install from the latest github version?

To confirm, if you run the following code, the output should be something like 0.0.4+51.gefef838 instead of just 0.0.4.

python -c 'import sys,neuromaps; print(sys.version); print(neuromaps.__version__)'