Kaustav Sen

2021-03-29


 

Who are you, and what do you do?

Hello! My name is Kaustav Sen (Twitter: @kustav_sen, GitHub: kaustavSen). I am currently working as an actuarial analyst in a consulting firm based out of Gurugram, India. I am specializing in the life insurance sector and alongside studying towards qualifying as a fellow actuary (hopefully by the end of this year!). Working in the actuarial field involves using mathematics, statistics and domain expertise to solve business problems facing an insurance company. This also means balancing the needs and expectations of both policyholders as well as shareholders amongst others.

How do you use R? What are your favorite packages?

I was first introduced to R as an undergraduate in college. However, it was more of a passing interest and I did not pick it up again properly until 2019.

This time our company had launched an internal course on predictive analytics using R. This was also the first time that I got introduced to the tidyverse ecosystem and immediately fell in love with it. I still remember finding the assignment operator (<-) and the pipe (%>%) somewhat awkward coming from a more traditional programming background in C and Python. Although overtime this has now become second nature and there is no turing back!

The predictive analytics course had piqued enough interest in me to continue to explore and learn more about R on my own. This led me to the R for Data Science book by Hadley Wickham and Garrett Grolemund. I also dicovered David Robinson’s screencasts which in turn introduced me to the #TidyTuesday challenge on Twitter. I started participating in this weekly challenge since mid-2020 which has helped me immensely in improving my R skills and also learn a bunch of new stuff!

My favorite packages apart from core tidyverse packages are: (particularly for visualizations)

  • ggforce for the amazing geoms it provides.
  • patchwork to effortlessly combine plots
  • ggfx a is relatively new package which allows you to apply filters and effects to normal ggplot layers
  • gt which provides a grammar for tables akin to what ggplot does for plots

What hardware do you run R on? Do you use any other software tools?

I have a couple of Dell laptops (one for work and one personal) both running Windows. Unfortunately, I haven’t had a chance to explore other operating systems although I have heard / read that Windows is the least friendly os when it comes to running R!

Our goto tool at work is still Excel (gasps) and a couple of other in-house developed software solutions. Other than that I occasionally also use Python.

Do you have any projects that you’re working on right now?

My main side-project at the moment are the weekly #TidyTuesday challenges. I am also planning to setup my own blog hopefully by the end of this year.

Other than this, I am keen to dabble into D3.js and creating some interactive visualizations. So, possibly something in this front most likely using insurance industry data over 2021!