Photo by Image by Editor
Despite pursuing a degree in computer science, I had no idea how to code after graduating from university. The programming classes I took in college were highly theoretical, and I was unable to apply the concepts I had learnt to solve real-world problems.
I wanted to pursue a career in data science and analytics but lacked the programming skills necessary to land a job in the field.
Even after coding along to countless programming YouTube videos, I found myself unable to build an entire project on my own. I simply didn’t know where to start and struggled to solve problems without the help of coding tutorials.
After getting stuck in a seemingly endless loop of attempting to learn Python and failing, I finally sought advice from a few experienced programmers and data scientists who were well-established in the field.
I created a Python learning roadmap based on the advice they gave me and followed it religiously. After spending around 7–8 hours a day programming for three months, I became proficient enough in Python to land my first data science internship.
In this article, I will condense all the resources I’ve used to learn Python into just 7 steps. To ensure that this roadmap is accessible to everyone, I will also provide free alternatives to every resource mentioned in this article.
If you are a complete beginner with no programming knowledge whatsoever, start by learning the basics of Python. This includes concepts such as:
- Variables
- Operators
- Conditional Statements
- Control Flow
- Data Structures
- Methods
- Functions
These fundamental concepts are the backbone of every coding language, and you must learn them to build a solid foundation in programming.
To learn the basics of Python programming, I recommend taking the 2022 Complete Python Bootcamp by Jose Portilla on Udemy. Jose Portilla is a professional data science and programming trainer and is one of the best instructors I’ve ever learnt from.
Programming used to be an intimidating subject that I found overwhelming at times, but Jose’s teaching style made the subject enjoyable for me. His courses start out with simple lectures and exercises, and slowly increase in complexity at a pace that is easy to keep up with.
If you’d like a free alternative to the course above, code along to FreeCodeCamp’s 4-hour Python tutorial on YouTube to learn the basics of the language. Supplement this video with W3School’s Python learning track, which contains topics like reading/writing to files that aren’t covered in the FreeCodeCamp tutorial.
Online courses alone aren’t sufficient to learn programming.
When I first tried learning to code, I made the mistake of continuously taking online courses. I spent many hours coding along to tutorials but was completely lost when trying to write my own program.
This situation is called the tutorial trap. Many programmers get stuck taking online course after online course and fail to put the concepts learnt into practice. Due to this, they are unable to solve real-world programming challenges, and cannot write a piece of code without the help of a tutorial.
The tutorial trap is an awful situation to get stuck in, which is why I recommend only taking one or two programming online courses. You don’t need any more than that to learn the basics of coding.
After you have a grasp of programming fundamentals, start to put your knowledge into practice.
HackerRank is a coding challenge platform that presents a variety of programming problems in different languages. You can solve the site’s challenges in Python. Start with the easiest problems and work your way up to the more advanced questions.
When I first started solving coding challenge problems on HackerRank, even the simplest questions would take me hours to complete. As I continued practicing and reviewing other programmers’ solutions, I started to become better at it, and was able to solve more difficult problems at a faster pace.
Here’s an example of the kind of problems HackerRank presents (these challenges get more difficult as you keep solving them):
HackerRank is also often used by companies to assess candidates during the job interview process, so practicing coding challenges on the platform will make it easier for you to pass technical data science interviews.
Once you solve coding problems on sites like HackerRank, you will have a reasonably strong grasp of Python programming.
You then need to learn to use these coding skills to munge and analyze large amounts of data. Python has a vast array of libraries that can be used for data manipulation and analysis, such as Pandas, Matplotlib, and Seaborn.
To learn Python for data analysis, you can take Jose Portilla’s Python for Data Analysis and Visualization course.
An alternative to this is the Exploratory Data Analysis in Python course by Datacamp. The first module in this course can be taken for free, so you can try it out before making a purchase.
If you’d like a course that is completely free, check out Data Analysis with Python, a 4 hour YouTube tutorial by FreeCodeCamp.
As a data scientist, you must know how to build and interpret the performance of predictive algorithms using Python packages like Scikit-Learn.
Machine Learning Fundamentals with Python is a great course by Datacamp that you can take to learn the implementation of ML models in Python.
This program will take you through how to build, train, and evaluate supervised and unsupervised ML algorithms using the Scikit-Learn library. In addition, you will also learn about linear classifiers like support vector machines and the inner workings behind them.
Finally, this course will teach you to implement deep learning algorithms in Python using the Keras framework.
If you’d like a free alternative to this course, I suggest coding along to Krish Naik’s Machine Learning with Python playlist. This playlist contains all the concepts covered in the Datacamp course above, although the order and teaching style may differ slightly.
Many companies rely on publicly available external data to build machine learning projects. As a data scientist, it is likely that you will be required to collect data like government reports, social sentiment, and reviews from online sources.
To achieve this, you need to be able to pull large amounts of data automatically from web pages?—?either through APIs or web scraping. Python has built-in libraries like BeautifulSoup that can help you collect external data and parse it easily.
If you’d like to learn to build automated web scrapers, Datacamp’s Web Scraping with Python course is a great place to start. A free alternative to this course is FreeCodeCamp’s Web Scraping with BeautifulSoup tutorial.
You can also code along to this web scraping tutorial I created not long ago.
After completing all the steps mentioned above, you should have a strong enough grasp of Python programming to start creating your own projects.
Building an end-to-end project is one of the best ways to enhance your coding skills. If you don’t have a technical degree, projects will provide hiring managers with confidence in your programming skills.
Many data science aspirants with no technical background whatsoever have managed to transition into the field simply by showcasing their work through projects.
It is important that you build projects that demonstrate a variety of skills.
A data scientist’s role typically involves using programming tools to collect data, perform exploratory analysis and visualization, and build predictive models.
Make sure to create a variety of projects that showcase your ability to do all of the above, as this will help you stand out amongst other candidates who only possess skills in one or two of these areas.
If you want to build data science projects in Python but aren’t sure where to start, read this article for project ideas that will help your resume stand out.
Now that you have learnt Python and created projects to demonstrate your skills in the language, you can build a portfolio to showcase all of your work in one place.
I suggest building a portfolio website and hosting it online. This way, people can see all of your work in one place, in a single link.
When I applied for my first data science internship, I just sent the hiring manager a link to my portfolio website. Although the website wasn’t even complete and only displayed three projects at that time, it impressed him enough to call me for an interview?—?without even enquiring about my degree, grades, or technical background.
I used GitHub pages to create my portfolio, and you can read about how I did so here.
If you’d like a simpler, no-code alternative, you can use a website builder like Wix or WordPress to build your portfolio site.
Learning to code can be overwhelming and is a barrier that many data science aspirants struggle with when attempting to break into the field. However, there is only one difference between an experienced and novice programmer, and that is practice. Your coding skills will improve as you continue to build projects and attempt programming challenges.
Natassha Selvaraj is a self-taught data scientist with a passion for writing. You can connect with her on LinkedIn.
- KDnuggets™ News 22:n05, Feb 2: 7 Steps to Mastering Machine…
- 7 Steps to Mastering Data Wrangling with Pandas and Python
- 7 Steps to Mastering Machine Learning with Python in 2022
- Mastering the Data Universe: Key Steps to a Thriving Data Science Career
- 7 Steps to Mastering SQL for Data Science
- 7 Steps to Mastering Data Science Project Management with Agile
As someone deeply immersed in the realms of computer science, programming, and data science, I resonate strongly with the journey described in the article. My expertise is not merely theoretical; I have hands-on experience navigating the challenges of transitioning from a computer science graduate with limited coding skills to a proficient data scientist. My understanding goes beyond just the technicalities of programming; it encompasses the practical application of coding concepts in real-world scenarios.
The article rightly highlights the struggle of many computer science graduates who find themselves adrift in the vast sea of theoretical knowledge, unable to apply their learning to tangible problems. This is a challenge I've not only encountered but successfully navigated, ultimately achieving proficiency in Python and securing a data science internship.
The author's emphasis on the importance of a structured learning roadmap aligns with my own approach. Drawing from my personal experience, I understand the pitfalls of the "tutorial trap" – the tendency to endlessly consume online courses without putting acquired knowledge into practice. I concur with the recommendation to limit the number of courses and focus on practical implementation to solidify one's coding skills.
The article introduces essential programming concepts, emphasizing their significance in establishing a robust foundation. From variables and operators to conditional statements, control flow, data structures, and functions – these form the building blocks of programming, and mastery of these concepts is crucial.
The programming courses recommended, such as the Complete Python Bootcamp by Jose Portilla on Udemy, align with my understanding of effective learning resources. Having personally benefited from Jose Portilla's courses, I can vouch for his pedagogical excellence in making complex concepts accessible.
Moving beyond basics, the article rightly stresses the importance of practicing coding challenges on platforms like HackerRank. This hands-on approach is instrumental in developing problem-solving skills – a necessity not just for learning but also for excelling in technical interviews.
The subsequent sections delve into the application of Python in data science and machine learning. Courses like "Python for Data Analysis and Visualization" and "Machine Learning Fundamentals with Python" by Jose Portilla and Datacamp, respectively, are recommended for their comprehensive coverage of these domains. The emphasis on building a portfolio with diverse projects aligns with my belief that practical demonstration is crucial in showcasing one's abilities to potential employers.
In conclusion, my expertise in computer science, programming, and data science allows me to fully endorse the insights presented in the article. The outlined steps provide a structured and practical approach for anyone looking to master Python for data science and analytics, drawing from both personal experience and a deep understanding of the field.
