Data Science Crash Course 10/10: Visualisation
This is 10th and the final instalment of Data Science Crash Course. I’m going to discuss how to present a Data Science project so that it’s appealing and instructive to others. In other words, let’s talk about visualisation.
How to visualise in Python
I start by writing that a well-documented Jupyter Notebook is perfect for a visualisation, especially if you want to show your work to someone with technical background. This is often what employers expect from you to send, when you do a test problem for a job interview.
Otherwise, if you’re talking with non-technical people, you will need more visual ways to show what you have achieved like:
- plot data using plotly or matplotlib
- create a heatmap
- show statistics in a table
Here’s an example of a simple plot using matplotlib:
The code for it is:
In general it’s amazing what you can create with plotly. Below you can find a sample of visualisations from plotly, all done in Python:
A good command of plotly will allow you to do really great things.
Going even further with plotly is Dash, by plotly, which allows you to build web-apps from Python. So it’s really great to show to anyone really, whether a technical or non-technical. You can have a look at what is possible with Dash here. For example this interactive dashboard was done in Dash:
As you can see, Python gives you amazing options for visualising and presenting your work. Now it’s your turn to play with it in Jupyter Notebook.
Share your open-source projects
The best way to share your projects is definitely Github. Upload your code with a well-written documentation and share it with your colleagues to get feedback. Of course you can’t really do it, if you’re working for a company, but if you’re working towards your own open-source project then Github is a perfect way to share your work. Data Science has a great community so you’re definitely going to hear feedback and learn more from others.
And that’s all for this Data Science Crash Course. I hope you’ve enjoyed it and learned useful things and you’re ready to tackle some data science problems now.
Let me know in the comments about your feedback and what else would you like to learn.
If you prefer a video version of this lecture, have a look here: