tknapen/nsd_access

python package to access the data of the NSD (natural scenes dataset) fMRI project

nsd_access's Issues

1 based indexing mismatch between shared1000 filenames and index in nsd_stimuli.hdf5

Hi Tomas,

I just realised that there is a mismatch between the filenames in

os.path.join(nsd_dir, 'Stimuli', 'shared1000', '*.png')

and the image brick.

so when i want to read the stimulus : shared0953_nsd69215.png if I follow the logic explained in the nds_dataset.doc, 0953 is the index in shared1000 space, and 69215 is the index in 73KIDs space.

However, if I want to read and show that same image from the brick, using nsda.read_images() i need to query for 69214.

were you aware of this? I'm wondering then if the same thing happens with the captions? i.e. if I want to return the caption for shared0953_nsd69215.png, should I query for 69214?

ian

error loading for volume based atlases Kastner2015 or thalamus atlas (func1pt8mm)

nsd_access/nsd_access/nsda.py

Line 220 in ffa837a

mapp_df = pd.read_csv(os.path.join(self.nsddata_folder, 'freesurfer', 'fsaverage',

@tknapen we have a bug with the mgz.ctab files not being read by pd.read_csv at the line referenced above. This works well for the nsdgeneral, visualsulc, or HCP_MMP1 atlases, but fails for Kastner2015 and thalmus.

Here is the full error trace (atlas=Kastner2015, sub=subj01, data_format='func1pt8mm'):

---------------------------------------------------------------------------
ParserError                               Traceback (most recent call last)
<ipython-input-30-9b391304df40> in <module>
      2 mmp1, atlas_mapping = nsda.read_atlas_results(subject=sub,
      3                                               atlas=atlas,
----> 4                                               data_format='func1pt8mm')

~/env/lib/python3.6/site-packages/nsd_access/nsda.py in read_atlas_results(self, subject, atlas, data_format)
    219 
    220         mapp_df = pd.read_csv(os.path.join(self.nsddata_folder, 'freesurfer', 'fsaverage',
--> 221                                            'label', f'{atlas_name}.mgz.ctab'), delimiter=' ', header=None, index_col=0)
    222         atlas_mapping = mapp_df.to_dict()[1]
    223         # dict((y,x) for x,y in atlas_mapping.iteritems())

~/env/lib/python3.6/site-packages/pandas/io/parsers.py in parser_f(filepath_or_buffer, sep, delimiter, header, names, index_col, usecols, squeeze, prefix, mangle_dupe_cols, dtype, engine, converters, true_values, false_values, skipinitialspace, skiprows, skipfooter, nrows, na_values, keep_default_na, na_filter, verbose, skip_blank_lines, parse_dates, infer_datetime_format, keep_date_col, date_parser, dayfirst, cache_dates, iterator, chunksize, compression, thousands, decimal, lineterminator, quotechar, quoting, doublequote, escapechar, comment, encoding, dialect, error_bad_lines, warn_bad_lines, delim_whitespace, low_memory, memory_map, float_precision)
    683         )
    684 
--> 685         return _read(filepath_or_buffer, kwds)
    686 
    687     parser_f.__name__ = name

~/env/lib/python3.6/site-packages/pandas/io/parsers.py in _read(filepath_or_buffer, kwds)
    461 
    462     try:
--> 463         data = parser.read(nrows)
    464     finally:
    465         parser.close()

~/env/lib/python3.6/site-packages/pandas/io/parsers.py in read(self, nrows)
   1152     def read(self, nrows=None):
   1153         nrows = _validate_integer("nrows", nrows)
-> 1154         ret = self._engine.read(nrows)
   1155 
   1156         # May alter columns / col_dict

~/env/lib/python3.6/site-packages/pandas/io/parsers.py in read(self, nrows)
   2057     def read(self, nrows=None):
   2058         try:
-> 2059             data = self._reader.read(nrows)
   2060         except StopIteration:
   2061             if self._first_chunk:

pandas/_libs/parsers.pyx in pandas._libs.parsers.TextReader.read()

pandas/_libs/parsers.pyx in pandas._libs.parsers.TextReader._read_low_memory()

pandas/_libs/parsers.pyx in pandas._libs.parsers.TextReader._read_rows()

pandas/_libs/parsers.pyx in pandas._libs.parsers.TextReader._tokenize_rows()

pandas/_libs/parsers.pyx in pandas._libs.parsers.raise_parser_error()

ParserError: Error tokenizing data. C error: Expected 2 fields in line 2, saw 6

option to have list of image_indices?

nsd_access/nsd_access/nsda.py

Line 333 in d0937b8

def read_image_coco_info(self, image_index, info_type='captions', show_annot=False, show_img=False):

when fetching the captions for a long list of indices, you end up having to loop through image_indices before calling read_image_coco_info for that current image in the loop, ending up in many i/o to load the captions from memory at each loop iteration.

could one pass a list of indices (or a single image_index) and return a list of captions (or a single caption)?

The usage of “json_normalize” has been changed in the latest pandas version

ImportError: cannot import name 'json_normalize' from 'pandas.io.json' (D:\anaconda3\envs\SD\Lib\site-packages\pandas\io\json_init_.py)

I use "pandas 2.1.1" in my project, and when I tried to from nsd_access import NSDAccess, I found that in the latest version, pandas used "pandas.json_normalize" instead of "pandas.io.json.json_normalize"

To fix this issue, replace from pandas.io.json import json_normalize with from pandas import json_normalize in nsda.py

Fix annotations reference frame

For the experiment, COCO images were cropped to be square. This means that the spatial parameters returned by the COCO interface are off, and need to be re-mapped to correct for the cropping.

set up setuptools installation

Need to add a valid setup.py installation route.