Altair Visualization
We will be using Altair in our course. It is a declarative visualization package in Python that is based on Vega-Lite which leverages the grammar of graphics.
Saving Altair Charts
Just installing altair and altair_saver will not allow you to leverage the .save() method to save your chart. The javascript visualization you see in your interactive python window needs additional external applications to allow .save('chart.png') to work.
We will go through a few ways for us to save our Altair plots.
1. Saving altair plots programmatically
Let’s say we want to save the above plot as a PNG file. Assuming we have already installed the altair library, we need to install the altair_saver.
1.1 Installing the altair_saver
Within your interactive python window execute the following command.
import sys
!{sys.executable} -m pip install altair_saver
1.2 Additional tool for saving plots
In the Github repository for altair_saver, the developers explicitly tell us to install additional tools. We suggest using NodeJS to get alrair_saver to work. However, you are more than welcome to study Selenium for further understanding.
NodeJS Installation
- Install the NodeJS for your platform
- Run the following in your Terminal (Mac) or Command Prompt (Windows) to install all the packages we need from NodeJS.
npm install -g vega-lite vega-cli canvas
1.3 Saving a plot using altair_saver
It might require you to restart VScode and import everything again for this to work. Please note that the plot will be saved in the same folder of the script.
chart = alt.chart(<data>).<chart_methods>
chart.save('name_of_chart.png')
2. Save as PNG method
The method only requires us to have Altair library. Whenever we output a plot, we will see a button with three dots at the top right corner of the plot.
Clicking Save as PNG will bring us to a window to save our plot.
3. Screenshot method
If all thing fails and we need to save a plot, the snip & sketch (Windows) or taking a screenshot (MacOS) will be our last resort.