search

Dlux804 / McQuade-Chem-ML

Development of easy to use and reproducible ML scripts for chemistry.
5 stars 1 forks source link

Moleculenet classification data sets not working properly. #60

Closed dickeygh closed 4 years ago

dickeygh commented 4 years ago

Describe the bug TypeError: object of type 'NoneType' has no len()error message is returned at line 56 features = pd.DataFrame(data, columns=columns) when trying to use both the clintox data set and the BBBP data set.

ValueError: Shape of passed values is (2, 1), indices imply (2, 201) error message is returned at line 56 features = pd.DataFrame(data, columns=columns) when trying to use both the HIV data set and the bace data set.

To Reproduce Steps to reproduce the behavior:

  1. Checkout the Feature_G branch from Github
  2. Run the different data sets by changing line 29 in main.py
  3. Running these data sets should reproduce the error.

Expected behavior These 4 data sets should be able to run successfully, similarly to the sider data set.

Screenshots Clintox error message BBBP error message HIV error message bace error message

Desktop (please complete the following information):

Other notes The error message seems to appear regardless of featurization option selected or model selected.

qle2 commented 4 years ago

Fixed in PR #70