Martin Ganchev, Instructor at 365 Data Science
Martin Ganchev is the author of the Python, SQL, and Integration courses in the 365 Data Science Program. He has an MSc in Economic and Social Sciences from Bocconi University in Milan, Italy. Throughout his studies, Martin gained advanced knowledge of Mathematics, Statistics, Econometrics, Time-Series, and Behavioral Economics & Finance. His experience includes assisting an empirical research for Innocenzo Gasparini Institute of Economic Research, and working for DG Justice and Consumers at the European Commission, where he handled data pre-processing, data quality checking, econometric and statistical analysis.
Hi Martin, could you briefly introduce yourself to our readers?
Hi! I’m Martin. I’m the author of the 365 Python and SQL courses, including SQL+Tableau, and Python+SQL+Tableau. My background is heavily related to quantitative disciplines such as mathematics, statistics, econometrics, and programming.
Right now, I am doing research for a new course, and I’m very excited. It will be about data preprocessing, data preparation, and data manipulation using Python. We’ll be dealing with the pandas and NumPy libraries, in particular. The course is going to be quite massive, with tons of exercises and hands-on tasks. Its content and structure will be unique and I’m really curious to see what its final shape will be like.
So are we! Martin, as you mentioned, you’re one of the 365 Data Science instructors who created our Python, SQL and Integration courses. In your opinion, which are the top 2 most in-demand programming languages in 2019 and what sets them apart from the rest?
Every programming language has its advantages and disadvantages. But more importantly, every language has its domain, and that’s where it can be better than other languages. For instance, some are specific to web-programming, others to data science, yet others have been created to manipulate relational databases, and so on.
What sets Python and Java apart is that they can operate in many domains, while at the same time they manage to preserve their integrity.
These domains are related to different fields of activity. Web development, machine learning, big data, statistics, and any sort of analytics intertwine and can all be part of a single business.
Another reason for you to start learning Python or Java is that they are quite intuitive to grasp.
And so on, and so forth… But as a whole, these are the reasons that make Python and Java really powerful today and I can’t see yet an alternative that would push any of them out of the top 2!
I can speak more about Python because that’s the one I’m specializing in. In fact, I’m still learning new things. And it is always great when you have to explore a new Python tool to obtain the desired analytic outcome!
Anyway, you must be aware which programming language will suit your needs best. In my case, I’m quite happy to say I use Python and SQL every day.
So Python and Java are at the top of your list. But which one is your favourite?
Python! Dare I answer differently?!
My main goal is to do analytics and Python gives me the opportunity to obtain the desired result in a quick and elegant way.
Its documentation is organized really well, its libraries are being updated daily… I love everything about using this language!
Moreover, I really enjoy discovering Python methods and tricks that others have already created to solve various issues, and then applying them while doing analysis or analytics. There is that feeling that you’re a part of a global community, but at the same time you can use Python to create something unique. It’s a really nice feeling.
Martin, you hold an MSc degree in Economic and Social Sciences, but you chose to transition into data science. Why did you decide to change your career path and how did you get interested in working with data in the first place? Any tips for people who want to successfully transition into a data science career?
I don’t see it as much of a change, to be honest…
After I had advanced in the fields Economics and Finance to a certain level (which, in my case, happened at some point in year 3 of university), I understood that what distinguishes an excellent analysis or academic paper from a poor one is not so much its topic or the econometric tools applied. It is the data that the researchers use to deliver their conclusions.
The most impressive papers I’ve read have been based on the researchers’ primary data that they had gathered in an outstanding fashion.
What I mean is, you can find or concoct great research ideas every minute, but they will be frivolous unless they are based on an amazing data set. Hence my love for preparing a high-quality data set!
There’s also another reason why I joined the world of data science.
It is growing at a huge pace, and the quantitative tools used in it are state-of-the-art.
How would I be able to adequately work in the field of economics in, say, 10-years- time, if I try to steer clear from data science today? The impact of the latter is immense, and it does affect the way we use economic and social sciences.
On a side note, I can say to those of you who are just about to transition into a data science career, regardless of your background, that it isn’t much different from any other scientific field. You will need to:
Read, apply, make mistakes, correct them, learn, and do this over and over again! On a loop! (wink-wink)
I suggest that you read about mathematics and statistics, business, programming, find connections to their previous knowledge and thus develop eclectic thinking. The latter is crucial. Moreover, it is very important to try to add style and vision to your coding and analyses. Otherwise, you won’t be any different from your peers.
Speaking of coding style, we recently asked our students on Instagram whether they’re early birds or night owls, and most of them chose the latter. Which type of coder are you?
I morph from one to the other. Particularly when a new course is in development (or a deadline is just around the corner!). That’s when I am at my best, and I can be both.