Guided Project: Popular Data Science Subjects on Stack Exchange

Hello all,

I am sharing my solution here. Please feel free to share your feedback.

GitHub Link

Mission Screen

script.ipynb (523.4 KB)

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

1 Like

Hi Ivan,

I’m doing the same project now. My approach is different from yours, but your project looks perfect anyway! It’s a great idea to query the DSSE yourself instead of using the available dataframe (well, I did the same :yum:). Perfect introduction to the problem, in-depth conclusion, project structure in general, clean code, great visualizations, interesting markdown observations. Also, very curious analysis about Python and R. This time the navigation works smoothly.

Here are some suggestions for your consideration:

  • Probably it’s better to remove from the introduction the details on how to fetch the number of sites on StackExchange: indeed, this information is merely mentioned on their home page :slightly_smiling_face:
  • The pieces of SQL queries: better to divide them into several lines for better readability, according to the SQL style guide.
  • The code cell [7]: you can combine these code lines in one.
  • The chapter “Analyzing the Deep Learning Trend”. You probably should mention before it, or right in the beginning of this chapter, why exactly we are interested in deep learning.
  • Some technical details in markdown (like those between the code cells [7]-[11], or between these ones[13]-[15]) can fit better as comments in the corresponding code cells.

Well, nothing else to add from my side. Awsome work, keep up this high level!

2 Likes

Hi Elena,

thanks for the feedback.

I kind of prefer 3 lines of code in cell 7 :slight_smile: Seems easier on the eyes.

GitHub has rendered my SQL code with proper spacing, which unfortunately cannot be said about NBViewer. I don’t know how to fix the NBViewer version as I believe that is not something I can generate myself. Do you know if I re-upload my code, will a new link be generated?

Oh, about SQL code and Nbviewer, then it seems that I’m going to have the same issue also with my project! :frowning: Next week I’ll finish mine and then will let you know how I would resolve it (well, if would resolve at all :sweat_smile:).

As for the code cell [7], I mean, combining the code, but then splitting it anyway into 3 rows. Each replace at its own row, then the code will be clean, readable and not repeating. But of course it’s not a big deal anyway :joy:

About generating a new link: no, it’s not updated after re-loading the notebook. But no worries, you can generate it yourself. Tomorrow I will send you the instructions from Sahil on how to do it. Otherwise, I can help you generating it, after that you reload your project, just let me know.

Hi Ivan,

As I promised, here are the instructions on how to re-render you notebook into nbviewer manually:

In very short words, you have to take this construction:
https://nbviewer.jupyter.org/urls/community.dataquest.io/uploads/short-url/
and then attach to its end (immediately, without any gaps) the name of your notebook. But, not the normal name which you gave to that file, but the subsequence of symbols automatically generated when you upload your notebook to your post. For example, after uploading your notebook, you have the following automatic notification in a post you’re going to share:

Untitled.ipynb|attachment](upload://wVLgAsnGlxmmjZ9gvvAAUlr7iM.ipynb) (306.6 KB)

Hence, you take this part wVLgAsnGlxmmjZ9gvvAAUlr7iM.ipynb (of course, in your case the subsequence of symbols will be different) and attach it to this:
https://nbviewer.jupyter.org/urls/community.dataquest.io/uploads/short-url/

Now, you can open the complete link in the browser:

https://nbviewer.jupyter.org/urls/community.dataquest.io/uploads/short-url/wVLgAsnGlxmmjZ9gvvAAUlr7iM.ipynb

and then you can share it as a link to your project in nbviewer here. The text of the link is usually this one:
Click here to view the jupyter notebook file in a new tab

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

1 Like

Elena,

it worked! I was able to fix the SQL string formatting by wrapping them in <pre></pre> tags. Thanks so much for showing me how to generate another NBViewer link.

1 Like

Ivan,

Thank you too for your suggestion! I will use these tags for my current project. Also, I’m glad that my instructions were helpful!

Congratulations on becoming a Community Champion! :trophy:

1 Like