Sankey Plots with Python
summary
Compares five Python approaches and shows what actually works on real Berlin census flows: raw matplotlib requires many manual tweaks but reproduces classic Sankey geometry, pythankey and apyankey accept dataframes quickly but limit customization, HoloViews + Bokeh gives interactivity with minimal code, and Plotly gives the most control at the cost of long layout fiddling. Finds an unexplained ~4,500 people "disappeared" in the data, demonstrates node repositioning hacks, and reports matplotlib: manual but faithful and plotly: fully adjustable.
AI-written summary · about
chapters (6)
related talks
01h 38mEffective Data Visualization
02h 23mEffective Data Visualization
01h 28mPython Visualization for Exploration of Data
01h 55mInteractive Data Visualizations with Python
02h 31mTech Tutorial: Data visualization in Python
02h 14mData Applications with Bokeh
35mVisualizing research data: Challenges of combining different datasources
30mVisualising data with Python