Python for Data Science –Free Learning Path
It took me a long time to learn Python. The reasons are aplenty on why it was such a long journey. However, hands down, I can tell you the main reasons are due to the vast amount of flexibility with Python. I tried to learn too much at one time instead of focusing on key areas of the language that are applicable to my life as a data analyst.
So the key for me was understanding that I wanted to focus on data science so I could ignore a vast amount of information and focus my time and effort on learning the skills and libraries applicable to data science. This would make the language invaluable to my day to day as an analyst. Here is the best pathway that allowed me to quickly accelerate my learning and apply my skills.
Phase I–Getting Your Feet Wet
Intro to Python.
For me, I found that this looking for intro classes are a major headache. It is very easy to get lost in learning all the ins and outs of a programming language when you just need the fundamentals for your data science and to move on to a project and go back to what you would need to learn to complete it. However, you need to get the foundations.
SENTEX Python 3 Tutorial Series:
Sentdex has been around for a while and I think he does a great job in explaining basics. However, I find his machine learning tutorial not as easy to follow but the basic tutorial will quickly allow you to get your feet wet.
Intro to Python Program for Data Science at DataCamp
For my purposes, Data Camp is awesome mainly because their courses are immediately applicable and this one is free. You can watch these videos at 2Xspeed and practice your coding within their environment.
A mobile app that helps you practice your skills
I was really skeptical about using a mobile app. But it turns out to be a great way to practice. So I would recommend getting the DataCamp app to reinforce what you learn, you can’t code in the app. It’s mostly a multiple choice skills test that reminds of what you’ve learned and prompt questions that you can then go and research and answers which helps you get a deeper grasp of the fundamentals.
Intro to Data Structures
Understanding data structures is super important. This foundational guide gives you a strong start with Pandas. You can get a start on understanding the nomenclature such as series, scalar, and data frames.
Pandas Intro Data Structures at Pydata.org(tutorial)
Microsofts at the end
After you finish the above resources, I would check out the course below by Microsoft. I find everything put out Microsoft very complicated for a beginner. So tackle this after you have built up your confidence
Microsoft Intro to Python for Data Science at EDX(full course)
This is a free course. If you want the certificate you need to upgrade it to $99. It’s not necessary
Phase II: Practicing Data Manipulation, Pandas and Following the Leader
Data Analysis in Python with Pandas(Video Course)
Any Python data science practitioner is going to use Pandas to manipulate data. Everything from loading in your data to pivot tables to creating filters can be done with Pandas. So get good at it!
The best series by far is by Kevin Markum at Data School. His delivery is super slow but very precise. I watch his vids at 2X and still find that super easy follow. I would recommend doing his whole Pandas series prior to moving on to the machine learning sections.
Kaggle is one the best websites around for a data science competition. However, it’s an excellent place to evaluate other peoples code from beginner to advanced. Check the kernels that have beginner attached to the title. These are great tutorials. Also, some of the best data sets can be found on Kaggle but you check out the guide to know what data set is best for different types of analysis.
Using Pandas for Better(and Worse) Data
Using Pandas for Better (or Worse) Data is a very comprehensive video on preprocessing and analyzing data. It helped me understand the idea of slicing and dicing data using pandas.
Dataframe Manipulation with Pandas(tutorial)
This course will get you ready for creating slices, filtering, and merging data frames. These all things that an analyst should be familiar with. So, the learning curve is relatively small since you can easily make the mental leap to with logic you have already acquired
Phase III. Your First Model
This was actually my ultimate goal was to be able to utilize a machine learning library. Scikit Learn is a great library to get your beak wet with machine learning. It’s fairly easy to use one of the best resources is again at Kaggle with a drum roll.
How Models Work by Dan B on Kaggle(tutorial)
Suraj Ruval has an amazing Youtube channel on Machine Learning and AI. You should definitely watch this for inspiration and all the amazing tips. But if you are at the beginning stages, it’s overwhelming. However, he has created a great 3-month curriculum on Machine Learning. It is very comprehensive so you will need to speed some time concentrating on it.
Learn Machine Learning in 3 Months by Suraj Raval(course at Github)
Other Great Resources
- How to Run a Linear Regression in SciKit Learn
- Choosing a Machine Learning Model (video)
- The principle of Machine Learning (course)
Great blog,Thanks for sharing the information