Unsure of what you need for each data science portfolio project you create?
What you need is a blueprint you can apply each time you create a new one.
As a matter of fact, there are 7 simple steps you can take for each of your projects that, when taken, can help attract employers to you.
Let’s take a look at each of the steps…
Seven Steps to a Winning Data Science Portfolio Project
There are 7 steps to take when creating each project you add to your portfolio:
- Pick an appropriate name
- Write a description
- Link to your website
- Add a bunch of topics
- Select your license
- Structure the code properly
- Add a README that details your project
Let’s take a look at each.
Step 1: Pick an Appropriate Name
The first thing someone will see is the name of your project.
Don’t go crazy or try to come up with clever names that don’t make sense.
Make your name straight forward and as descriptive as possible.
More than likely you’re going to be sending people directly to your GitHub profile, so having a project name that’s memorable isn’t an issue.
The most important thing is to make it readable and obvious.
Step 2: Write a Description
The description is the second most important part of your portfolio project. It shows on your GitHub profile and on the project page.
Again, don’t try to be clever.
Provide a straight-forward description of what’s in the project.
Now, you don’t need too many details – that’s what the ReadMe is for.
What you do want is enough information to get someone interested in the project.
Step 3: Link To Your Website
Not much to say here other than always provide a link back to your personal website.
If you wrote a blog post about this particular project link to that instead.
Step 4: Add a Bunch of Topics
Topics do two things for you:
- They help other people find your projects based on their interests
- They further describe what your project is all about
Add topics for everything your project entails for maximum findability.
Step 5: Select a License
The license may seem like a minor addition to your project, however it is important.
Which license you choose is based on how open you want your code to be.
GitHub created an open source license selector to help you pick the right one.
Step 6: Structure The Code Properly
Your code is what employers are going to be digging into to see if you know what you’re doing.
Before looking at individual code files employers will see your project structure. Be sure to have straight-forward names on your folders, and break your code into separate folders only as necessary.
Super important is this – use Python naming conventions. That means lowercasing all names and separating words with underscores. This shows attention to detail and some knowledge of Python.
Additionally, be sure you have the following files:
- .gitignore – ignores files that don’t need to be a part of the project
- LICENSE – the license file mentioned above
- README.md – detailed below
- requirements.txt – a list of the non-core Python libraries required by your project, and their version numbers
If you create your repository from the GitHub website, you’ll have the first three files automatically.
Step 7: Add a README That Details Your Project
Always using an uppercased name for the file, the README is the file contains all the details of the project.
At a minimum you want a few sentences which tell what your project is all about.
Other helpful information you want to consider adding include:
- Information on getting set up to run your code
- A description of what’s in each folder
- Directions or links to directions for setting up anything your project requires
- Output screenshots – these can be helpful if your code produces something and let’s people know if they have been successful or not.
If you’re looking to get into the data science industry, building your data science portfolio is one of the steps. Find out what else to do with this free ebook.