How/What Correlation for Position Ranking?
-
16-10-2019 - |
Question
I'm newbie to Data Science. I'm trying to understand how to correlate the position of an app in the app store, e.g. 1-10, to the number of backlinks, e.g. #1 = 250 links, #8 = 50 links, to the app store listing page in Python.
I've manage to correlate both rows entirely, which obviously gives me a nonsense figure.
Could someone point me in the right direction to allow me to fill this knowledge gap. I'm not entirely sure of the correct terminology for this, making Googling it pretty impossible.
Solution
Person correlation assumes data is coming from a Normal distribution and there is a linear relationship. An alternative is the Spearman correlation or Kendall's tau for ranked data.
As an edit, here are the links to how you would calculate the Spearman Correlation coefficient and Kendall's Tau, respectively.
http://docs.scipy.org/doc/scipy-0.15.1/reference/generated/scipy.stats.spearmanr.html
http://docs.scipy.org/doc/scipy-0.15.1/reference/generated/scipy.stats.kendalltau.html
Best of luck.
OTHER TIPS
Sounds like you are looking for the Pearson correlation coefficient between the two variables. You can compute that by using Scipy's pearsonr method.