Blue Week Special Offer | Brighten your week!

Winning Jeopardy, but faster. Feedback welcome

Hello Dataquest community!

I’m sharing with you my solution to the Winning Jeopardy project. I’d like some feedback on it! Anything goes: code, visualization, narration, mistakes…

I took the extra step of making the code faster by:

  1. Computing the high/low value occurrences of words all at once, instead of looping over the database once for each word, as the guide suggests.
  2. Using a dictionary to store p-values and avoid recomputing them (i.e. memoization). I’m not sure this second step is very useful, but it was fun implementing it!

With that, one can actually compute all needed p-values in 3 seconds. Some of the words do have very low p-values.
Winning Jeopardy.ipynb (144.8 KB)

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