Following our Udemy tutorial, we realized that that was only one way to go. As soon as we hit some small problems, things started to get complicated. Online tutorials give you a dense but superficial overview, where everything from code to dataset is neatly prepared. But real word stuff is messy rather than neat.
We were frustrated with a couple of things, that’s why we looked for some additional resources. We used a new visualization library and pandas-profiling for our problem with missing data.
We wanted to explore different ways of visualization and found seaborn.
Seaborn is a Python visualization library based on matplotlib. It provides a high-level interface for drawing pretty nice statistical graphics. It’s easy to install and implement. Check out the example gallery and the seaborn tutorial.
The online course we chose did not introduce seaborn yet, but hopefully it will come up at some point.
We had a tough time analyzing a dataset with a lot of missing data, so we needed to find a way that gives us a nice and quick overview of what’s happening. One of our team members found this awesome .ipynb for data-profiling in Jupyter. It gives a great dataset-overview at a glance with a couple of lines of code. Definitely useful and worth checking out.
We’ll be posting more additional resources that we find. Until then, quote Enjoy Machine Learning! unquote 😀