kulbachcedric
commented
2 weeks ago Hi @atanikan,
thank you for your input! That helps a lot!
Hello! As a reviewer for JOSS, I’d like to suggest a few updates and additions to the software installation process. For simplicity, I’ve consolidated them here in a single issue.
- Python Environment Setup: It would be helpful to specify the minimum required Python version and include steps for setting up either a virtual environment or a conda environment for package installation. This will make the installation process smoother for users.
Good point I'll add it as tags to the README.md
- Link to Examples: I recommend adding a link to the examples folder in your README or documentation. This can help users quickly locate example scripts and adapt or modify them as needed.
Will be done!
- Jupyter Dependency: Since the examples utilize Jupyter notebooks, it would be beneficial to list jupyter as a dependency when installing the package via pip. I noticed it’s missing from the pip list ((deep-river-joss) ADITYAs-MacBook-Pro-2:classification adityatanikanti$ pip list | grep "jupyter").
I added Jupyter to the dev installation instructions pip install deep-river[dev]. As it is only required when developing and not when using the package.
- Broken Example: The example for the Simple Classification Model in this notebook isn’t working, as well as some other examples. The error output is as follows:
Copy code 0it [00:00, ?it/s] --------------------------------------------------------------------------- AttributeError Traceback (most recent call last) Cell In[19], line 6 4 metric = metrics.Accuracy() 5 for x, y in tqdm(dataset.take(5000)): ----> 6 y_pred = model_pipeline.predict_one(x) # make a prediction 7 metric = metric.update(y, y_pred) # update the metric 8 model_pipeline = model_pipeline.learn_one(x, y) # make the model learn AttributeError: 'NoneType' object has no attribute 'predict_one'
I'll revise this one :-)
Please let me know if any further clarification is needed on these points. Thank you!
I'll work on these points in the upcoming week and keep you up to date. Thanks again for you feedback!
Hello! As a reviewer for JOSS, I’d like to suggest a few updates and additions to the software installation process. For simplicity, I’ve consolidated them here in a single issue.
Python Environment Setup: It would be helpful to specify the minimum required Python version and include steps for setting up either a virtual environment or a conda environment for package installation. This will make the installation process smoother for users.
Link to Examples: I recommend adding a link to the examples folder in your README or documentation. This can help users quickly locate example scripts and adapt or modify them as needed.
Jupyter Dependency: Since the examples utilize Jupyter notebooks, it would be beneficial to list jupyter as a dependency when installing the package via pip. I noticed it’s missing from the pip list ((deep-river-joss) ADITYAs-MacBook-Pro-2:classification adityatanikanti$ pip list | grep "jupyter").
Broken Example: The example for the Simple Classification Model in this notebook isn’t working, as well as some other examples. The error output is as follows:
Please let me know if any further clarification is needed on these points. Thank you!