Karthik Vetrivel ('21)

Karthik

Analyzing COVID-19 with Python

Karthik is not yet an alum, but his work as a student is so relevant to our world right now, it made sense to share this with our alumni community.

Christian Fulghum (’77)

How are you doing at home on COVID-19 day whatever this is?

Karthik Vetrivel (’21):

Well, it's been good. Honestly, school's been about the same difficulty-wise for me but there has been a lot more time to just go into any hobbies I have. People have been saying COVID is a really good time to do something you haven't had the time to do before. So just getting a chance to catch up on exercise and any other hobbies that I’m interested in. Like skateboarding, I've been getting into that recently, which has been really fun.

Christian:

Cool. Be careful out there. The reason we're talking today is because you responded to a shout-out that I did to the alumni community asking, "is anybody working on anything scientific or healthcare-wise related to COVID-19 specifically?" And to my surprise, you a current student, responded. I guess I shouldn’t be surprised after interviewing Ashwin Appiah (’06), who specifically mentioned that your generation is way ahead in developing its technological skill set. Tell me a little bit about what you're doing.

Karthik:

This summer I actually began to look into how you can use certain programming languages like Python to visualize and analyze data. I learned recently from my sister that the New York Times and John Hopkins CSSE publishes a lot of data, specifically at the county-level, on how COVID-19 is progressing. They have the number of cases and deaths, but I could also find things like population density online. And using that I've been creating a ton of graphs and other statistical analysis on how population density might correlate with the number of cases per county. I wanted to analyze how COVID-19 was developing within King County and within Washington counties to understand its effects locally. Most people’s attention is focused on national statistics, but I think it is more helpful if we look at how it's growing in the areas right where we are.

Christian:

So basically you're using publicly available data and crunching the numbers, then creating visual representations?

 

Karthik:

Yes!

Christian:

How does Python help you analyze the data? I mean, are you basically entering a bunch of numbers into a database and then turn Python loose on it? Help me understand the method.

Karthik:

Back in the ‘90s there was actually this programming language called MATLAB, which people use a lot for scientific processing. So if you ever read any kind of scientific paper, you'll see graphs generated. In those papers, those are typically generated with MATLAB. So a lot of the graphs that you can generate in Excel or other graphing programs like that you can generate with Python and with Python you can do a lot of interesting things.

Python doesn't automatically do it. You have to specify what you want it to do. But you can create a lot of interesting graphs, especially interactive graphs that you can't do with Excel. Python is specifically great because it has a lot of libraries that make it easier to write the code.

For example, if I wanted to create an interactive graph that graphs the mortality rate of COVID-19 over time, I could easily do that in four lines of code, which isn't a lot for reference and Python just makes it really easy to do. Of course it still takes effort. I still have to do some data processing myself, which Python makes really easier, and I have to come up with an approach based on the data.

Christian:

And what made you pick Python as a language to get to know more about?

Karthik:

I was actually already familiar with Python. This summer I interned at this company called Foundry10, which is an educational research company in Seattle, and I wrote predictive analytical software using Python there. So they kind of taught me the ropes of how to use Python for analytical software and I transferred the skills over to using it on COVID-19 datasets.

Christian:

I understand the words predictive and analytical, but how might you apply that? Give me an example.

Karthik:

You might apply that to predicting how COVID-19 would spread in the future. Of course, with something like COVID-19 that's really difficult and a ton of people are trying to make accurate predictions on this. But for example, with Python you could compare COVID-19’s growth with how a virus like H1N1 has grown or MERS or SARS, because there are datasets available on those. You could see how those grow and then you could see how COVID-19 is growing so far and then compare the two growth patterns and see if there's any similarities between the two. So if you notice, “Oh, there is a similarity,” then you can start to make some inferences based on past data and see what might happen in the future.

Christian:

I see. And human behavior is baked into those numbers, I presume. In other words, we can assume that people have responded to those prior three outbreaks very similarly. One would think that being able to look at those prior large datasets would at least help to gain some understanding. Do you have any notion yet about how you'd like to apply these kinds of skills beyond high school?

Karthik:

Oh yeah, for sure. As I mentioned, I interned at an educational research company. So I've been really interested in how we can use this kind of software, specifically like computer science. And it's like this paradigm of machine learning to education. Because a lot of education is uncertainty and inequity. And I believe that computer science can help a lot of that.

One of the things that I worked on over the summer was a system for identifying at risk students earlier using predictive analytical software. I've already started, but definitely after college I want to use computer science in the educational context.

Christian:

So you're more interested in a research or educational context, say then on building a product that you sell or something that's more just purely for entertainment sake?

Karthik:

Right. Of course, it's a little too early for me to decide specifically what path I want to go on, but I'm just letting my interests guide me at the moment. I think that's what I'm interested in at the moment.

Christian:

And are there any coders in your family?

Karthik:

Oh yeah. My dad and my sister actually. My dad came to the US to study computer science for his masters at the University of Iowa. Then went to Microsoft. And then my sister (Sakthi Vetrivel (’15)) recently graduated from Caltech and she's working at Microsoft again.

Christian:

And she also participated in Making a Life Day this year, which is really cool. Does your dad still work at Microsoft?

Karthik:

No, he now works on his own venture, which works specifically on mentoring software for businesses. So it's still that paradigm with computer science, but he no longer specifically codes, he does more on the business side. But he misses coding because it can be fun. He talks about it often how he misses coding and working on projects..

Christian:

I really appreciate your taking time with me to share this. As an alum myself, your generation fills me with hope because you are so far along in your skills and your awareness as a global citizen compared with my generation. Keep up the good work, and stay in touch!

 

Karthik:

Thank you, I will!