First Review

[Tamoghna-Saha]

Bugs:
Environment: Virtual Env in WSL 2

• While installing: pyproject.toml file format issue.
• While launching .py script: ModuleNotFoundError: No module named 'requests'
• While launching .py script: Unable to access localhost:5789 from browser

Operations Performed:

• Removing duplicates
• Formatting columns value that was suggested
• Custom code implementation - typecast values in certain columns using regex

Scope of Improvements:

• Option to edit custom code selected from New Actions
• Error info displaying in the UI is not redirecting to the root-cause. Was able to figure out from logs.
• Info icon which on hover can explain some terminology like category_high_cardinality
• Similar info icon or hyperlink that redirects to what certain outliers means when selecting, specially how auto is performed
• Option to download the processed data in CSV/Excel format, currently only JSON format

[tommydangerous]

Thank you @Tamoghna-Saha so much for the feedback!

[tommydangerous]

@Tamoghna-Saha will definitely review and incorporate the feedback in the tool.

In the meantime, would you please join our online Slack community so we can keep you up to date? https://www.mage.ai/chat

Thank you!

[tommydangerous]

@Tamoghna-Saha for the 3 Bugs:

• While installing: pyproject.toml file format issue.
• While launching .py script: ModuleNotFoundError: No module named 'requests'
• While launching .py script: Unable to access localhost:5789 from browser

Was this caused by running pip install mage-ai?

[Tamoghna-Saha]

Initially, I was trying to create venv using the toml but it wasn't working.
The next thing happened after I was trying pip install mage-ai in a clean venv.

[tommydangerous]

Hey @Tamoghna-Saha , so sorry for the super long delay. We’ve been hard at work improving the tool.

We’ve actually changed it quite drastically to now support writing your own custom code in the tool.

With Mage, you can now transform your data and build production ready pipelines for ML.

Please take a look and go through the setup. If you have any issues, please ping me here, email me at tommy@mage.ai, or please join our Slack and chat with me there: https://www.mage.ai/chat.