README.md

Libraries used

Pandas : To process data
Numpy : For mathematical calculations
Matplotlib : Simple Data Visualization
Seaborn : To see trend between CodeForces rating and CodeChef rating

Why :

BitByte (Programming Club, PDPM IIITDM Jabalpur) is organising Most Improved Competitive Programmer
In which members are ranked based on improvement in rating on different platforms.
Comparing users who are active only on one platform was in-accurate simple sorting, hence CC to CF rating converter modelled using Linear Regression was developed.

How

1. Scraped data of around 12000 from StopStalk.
2. Solved data-discrepancies using Excel and Pandas
3. Spotting a linear trend between CodeChef and Codeforces rating (using scatter plot)

   

4. Parameters for best fit line were calculated.

   

Observations :

1. After initial data cleaning, we had users with non-nan 10280 CodeChef and 7437 Codeforces ratings .
2. 6442 Users had both CF and CC ratings.
3. Estimated coefficients for line y=mx+c are as follows :
   1. constant (c) = -32.62926507315615
   2. slope of line (m) = 0.8315620555789324
   3. Where y represents CodeForces rating and x represents CodeChef rating

Further development :

1. Update the dataset regularly over the course of MICP
2. Develop a web-app for the same