The talk covers how you can visualize your Google location history using Streamlit. It covers how you can get the data, cleanse it, augment it using images and finally visualize it using Python without using any Javascript libraries. This talk was given at EuroPython 2021 (https://ep2021.europython.eu/talks/4NTyz92-virtual-tourism-in-covid-times/)
7. Google Location Data
● Standardize
○ Timestamp in milliseconds
■ Convert to timestamp using
pd.to_datetime()
○ Latitude & Longitudes
■ Divide by 10^7
10. Jupyter Notebooks
● Great for Experimentation
● Graphs & Visualizations
● Storytelling using Data
● Customize with Code
● “I don’t like notebooks” - Joel Grus
11. Web Apps: Wild West of Javascript
● Lots of Frameworks /
Libraries
● Paradox of Choice
12. Streamlit
● Turn Data Scripts into Web
Apps
● Interactive
● No Frontend Experience
Required
● Easy to Deploy
17. Deployment
Private Server
● Config File
○ $CWD/.streamlit/config.toml
● Run
○ streamlit run script.py
Streamlit Sharing
● Public Github Repos for Free
● Premium Offering
18. Streamlit Observations
Pros
● Make Python Code Interactive
● Customizable
● Deployment is easy
● Active Community
Cons
● No built in
Authentication
● Limited Design
flexibility
Good to have for Data Scientists