Guided Project: Popular Data Science Questions!

Hey All,

I am sharing my latest project for any help or guidance. I feel that there were times during the project that I could have used simpler methods to solve the problems so if that is true please let me know.

Thanks a lot!

Popular Data Science Questions.ipynb (116.5 KB)

Click here to view the jupyter notebook file in a new tab

1 Like

Hi @jmerrell247,

Thanks for sharing your project with the Community! It looks nice, well-structured, with clean and well-commented code and good markdown explanations.

Here are some suggestions for your consideration:

  • It’s better to use “we” instead of “I” throughout the project.
  • I would re-phrase a little bit the introduction, since the context suggested by the DQ instructions (that we are working as data analysts in that company, etc.), is indeed hypothetical.
  • You can mention that you decided to use the Posts table from the database, since there are plenty of others.
  • Some obvious code comments (like # Import pandas, # View first 5 rows) can be omitted.
  • About the visualizations. It’s better to despite them, make titles bigger, remove the legend if it only contains 1 item (like in the code cells [10] and [11]). Also, a good idea is to create horizzontal bar plots instead of vertical, because here we have a lot of bars, and using horizontal bar plots would improve their readability. Finally, you might consider defining a function for creating these plots (since they are all created according to the same scheme, you won’t have to repeat the same piece of code).
  • Probably, you can add a couple of words about why we became interested exactly in deep learning and not in the super-popular machine learning.

That’s all from my side, hope it was helpful! I saw that you shared a lot of other projects recently, but for now I myself still haven’t done them, so my suggestions there would probably be less interesting :joy:

Happy coding!

Thanks for the tips Elena!

Will definitely take your points on board.

1 Like