The dynamicroutingtask from samgale

dynamicroutingtask's People

Contributors

Watchers

dynamicroutingtask's Issues

`subjectName` from hdf5 path vs. contents

Hey Sam,

Would it be a problem to revert back to the commented-out line for extracting subjectName below?

   class DynRoutData:

    ...

    def loadBehavData(self,filePath):

        self.behavDataPath = filePath
        
        with h5py.File(self.behavDataPath,'r') as d:
        
            # self.subjectName = d['subjectName'][()]
            self.subjectName = re.search('.*_([0-9]{6})_',os.path.basename(self.behavDataPath)).group(1)

        ...

I'm trying to use the class with hdf5s on s3 as follows:

import io
import upath     # 3rd party package for opening files from cloud storage

obj = DynRoutData()
file = io.BytesIO(
    upath.UPath(
        's3://aind-ephys-data/ecephys_670248_2023-08-03_12-04-15/behavior/DynamicRouting1_670248_20230803_123154.hdf5'
        ).read_bytes()
    )
obj.loadBehavData(file)

Which works fine for opening via h5py.File(), but can't be used as a path to extract the subject id.

If it's preferable to get the subject from the filename, perhaps we could do something like this:

    def loadBehavData(self,filePath, fileBytes=None):

        self.behavDataPath = filePath
        
        with h5py.File(fileBytes or self.behavDataPath,'r') as d:
        
            # self.subjectName = d['subjectName'][()]
            self.subjectName = re.search('.*_([0-9]{6})_',os.path.basename(self.behavDataPath)).group(1)

    ...