Kepler.gl is a powerful open-source geospatial visualization tool developed by Uber. It enables users to effortlessly create custom, interactive maps from large-scale location data. With its intuitive interface and extensive customization options, Kepler.gl empowers users to explore and analyze geographic patterns, trends, and relationships. Whether for urban planning, logistics optimization, or data journalism, Kepler.gl provides a user-friendly solution for visualizing complex spatial datasets. Its versatility, ease of use, and ability to handle big data make it an essential tool for anyone needing to understand and communicate insights from geospatial information effectively.
The following three libraries - geemap, leafmap, and streamlit-kepler - currently offer support for integrating Kepler.gl into Streamlit, facilitating data presentation. However, they are limited to static Kepler maps. "Static" implies that once the map is rendered, Streamlit cannot manipulate it further or retrieve its current state. To modify data, the map must be re-rendered. This restriction significantly constrains Kepler's utility in Streamlit. Many applications demand dynamic capabilities, enabling Streamlit and Kepler to communicate bidirectionally - adding or removing data at any time and responding to user interactions on the map promptly and flexibly.
We've created a Streamlit component for Kepler, enabling seamless bidirectional communication between Streamlit and Kepler. Now, Streamlit can effortlessly retrieve the map's current state and dynamically adjust the application accordingly. Additionally, Streamlit gains the ability to add data to the map on-the-fly, independent of whether the map has been previously rendered. This breakthrough empowers users to interact fluidly with the map, enhancing the application's versatility and responsiveness.
See demo at https://kepler.streamlit.app/
pip install streamlit-kepler-componentIn the present prototype, Kepler.gl accepts the following parameters:
- A list of GeoDataframes or Dataframes
- options
- config
- height
When there's an interaction with the map, such as a mouse movement, the Kepler streamlit component returns the current map configuration. In simpler terms, any action performed on the map will prompt the entire page to be re-rendered. To ensure smooth functioning and avoid potential bugs, it's essential to utilize session state management for your web page and execute any prolonged actions within a separate thread.
import json
import streamlit as st
import geopandas as gpd
from keplergl import keplergl
if "datasets" not in st.session_state:
st.session_state.datasets = []
sf_zip_geo_gdf = gpd.read_file("sf_zip_geo.geojson")
sf_zip_geo_gdf.label = "SF Zip Geo"
sf_zip_geo_gdf.id = "sf-zip-geo"
st.session_state.datasets.append(sf_zip_geo_gdf)
h3_hex_id_df = pd.read_csv("keplergl/h3_data.csv")
h3_hex_id_df.label = "H3 Hexagons V2"
h3_hex_id_df.id = "h3-hex-id"
st.session_state.datasets.append(h3_hex_id_df)
map_config = keplergl(st.session_state.datasets, height=400)