Learn how to create effective data visualizations using Python libraries Matplotlib and Bokeh in this hands-on training video from ODSC West 2019. Explore different types of visualizations, techniques for reducing clutter, and methods for guiding the viewer's eye. Discover when and how to add interactivity with Bokeh, and gain insights into visual design principles that enhance communication of data insights. Suitable for beginners and experienced data scientists alike, this 36-minute talk covers the importance of visualization in the data science workflow, common encoding techniques, and best practices for creating clear and compelling visuals to effectively convey information to various audiences.
Data Visualization - From Square One to Interactivity
Overview
Syllabus
Intro
Visualization in the data science workflow
Where we are today
Why make data visualizations?
Ask whether a visualization is necessary and effective
Visualizations can fundamentally surprise us
Visualizations are mappings, encoded and decoded
Types of data and attributes
Some quantitative encodings are easier to decode
Common versus noncommon scales
Faceting: four variables with position
Preattentive attributes are easily perceived...
Example: bar chart
References and further reading (all on Amazon)
Taught by
Open Data Science
