Unable to run .get_label_meaning() for the TWOSIDES dataset

[sbonner0]

Hi,

When I try to call the function .get_label_meaning() on the TWOSIDES dataset I get the following error:

python3.8/site-packages/tdc/base_dataset.py in get_label_meaning(self, output_format)
    163             dict/pd.DataFrame/np.array: when output_format is dict/df/array
    164         """
--> 165         return utils.get_label_map(self.name, self.path, self.target,
    166                                    file_format=self.file_format,
    167                                    output_format=output_format)

AttributeError: 'DDI' object has no attribute 'target'

I am trying to understand what Y represents for this dataset? I assume the ints are mappings to reported side effects but it would be great to get that clarified.

Environment:

• OS: MacOS 11.6
• Python version: 3.8.12
• TDC version: 0.3.1

[kexinhuang12345]

Hi, thanks for pointing it out. For now, you can get TWOSIDES side effect label by

from tdc.utils import get_label_map
get_label_map(name = 'TWOSIDES', task = 'DDI', output_format = 'df')

You can see a column named side effect name.
we will add support to it in the next release

[sbonner0]

Thanks @kexinhuang12345!

[kexinhuang12345]

Hi, you can now retrieve the label map via:

from tdc.utils import get_label_map
get_label_map(name = 'TWOSIDES', task = 'DDI', name_column = 'Side Effect Name')