4 Things to Know Before Your First Data Science Project

Introduction

A position in data science is the most desired job in the world right now. Just on LinkedIn alone, you will find almost 200K open data scientist positions around the world.

The statistics also predict the rapid growth of jobs in data science. According to the BLS, the demand for data scientists and other professionals in this field will keep increasing by 15% each year until 2029, with an annual salary of $122K on average. 

There’s no doubt that if you choose data science as a job, you will secure yourself an awesome future full of interesting challenges. But if you’re just a beginner, this job can seem quite overwhelming, especially after reading the requirements and responsibilities posted by the companies in data science job descriptions. 

Luckily, there’s no need to apply for a job or an internship right away to gain some skills if you want to become a data scientist. It’s absolutely OK to start a project by yourself and build up your knowledge until you’re confident enough. 

That being said, starting your own project is not as easy as it seems as well, especially if you don’t have any guidance. So, today, I’ll share with you some of the tips I learned the hard way when I was working on my first data science project and that I wish someone told me when I was just a beginner. 

1. Be Ready For Failure

I know this phrase may sound weird. Should you set up your mind for failure right away? 

Absolutely not. But sometimes, people who start in data science have an overly idealistic attitude and expect quick results from their work. And when they encounter their first obstacle, their spirits sink. 

So, my first advice would be to embrace failure and learn from it. I know it sounds cheesy, but I wish I adopted this mindset when I started my first data science project. Such an attitude would have saved me so much time I spent worrying. 

If you find it hard to cope with your mistakes, try to approach them with curiosity rather than desperation. Why is code failing? Maybe you’ve misused expressions as defaults in function arguments? Or is it a problem with class variables? Explore your mistakes, but don’t fret over them. 

2. Seek To Solve the Problems You’re Passionate About

The best thing about working on your own data science project is that you get to choose to work on whatever you want. But, at the same time, it can be a trap. 

It might be very tempting to take one case from a certain company, like Uber’s pickup analysis or predicting a café’s success, and try to solve this case by yourself. Here, the issue is that someone has already solved these problems, and you might be tempted to check in with their results just to see if you’re doing everything right. 

And, the second issue is—are you really passionate about these problems? Your first data science project should motivate you to advance in your knowledge, and if you pick the topic that doesn’t interest you, it’s unlikely that you’ll feel motivated. 

What if your passion is too “boring?" 

It doesn’t matter as long as you are into it. You can be passionate about zoos, so what’s stopping you from developing a database containing the nutrition facts and dietary directions every animal needs to remain healthy?

3. Don’t Try to Master All Programming Languages At Once

When I first started my career in data science, I was a blind perfectionist. I thought if I knew Python, R, JavaScript, and SQL all at once, I would be the most outstanding professional in the world. I believed in this idea up until my first boss said they were only interested in Python. 

The thing is that different projects require different programming languages. Ken Jee shares the same thought in his How I Would Learn Data Science video:

Basically, his idea is that Python rules the majority of data science projects, while R is used in academic circles but is also very common. This idea leads me to the point that you should research projects similar to the one you’re about to start, see which programming language is used, and then start improving your knowledge of it. 

The beauty of data science is that you’re always learning, non-stop. You’ll have time to learn another programming language if you’ll need it. But right now, just start small and keep your expectations realistic. 

4. Gain a Basic Understanding of Statistics

As you already know, it’s better to connect your first data science project with a real-life problem you’re passionate about. But the knowledge of a programming language is not the only thing you’ll need for that. 

While working on my first project, I found that the knowledge of statistics can come in handy as well. If you have a basic understanding of statistical analysis, you’ll be able to track meaningful trends and derive them from data with mathematical calculations. 

So, you might find it useful to pass a quick statistics course just to grasp the general idea of its role and how it can be applied in data analysis. 

What Else Should You Know?

In conclusion, I want to add one more thing to all the above-mentioned tips—keep practicing every day. In the video mentioned above, Ken Jee says that you should split your time 50/50 working on your own project and studying the code of others. 

That’s why resources like Kaggle are so popular among data scientists—you get a chance to practice your knowledge, ask questions, and get guidance. But it’s important to do it on a daily basis if you want to advance fast in data science. 

Hopefully, my tips have motivated you to start your first data science project. Now, get to work!