8 Visualizing location data with Datashader
This chapter covers
- Using Datashader to visualize many datapoints when downsampling isn’t appropriate
- Plotting interactive heatmaps using Datashader and Bokeh
In the previous chapter, we looked at a few ways to gain insights from data using visualization. However, every method we looked at relied on finding workarounds to reduce the size of the data we used for plotting. Whether by randomly sampling, filtering, or aggregating the data, we used these downsampling techniques to overcome the inherent limitations of Seaborn and Matplotlib. Although we’ve shown that these techniques can be useful, downsampling can cause us to miss patterns in the data because we’re throwing data away. This issue becomes the most evident when dealing with high-dimensional data, such as location.