andeplane-reveal-streamlit-component

Cognite 3D Reveal Streamlit

This is a Streamlit library that can be used to show 3D models inside a Streamlit app. It works especially good when used inside Streamlit in Fusion, but can also be used in standalone Streamlit apps.

How to install

You simply install it by running pip install cognite-streamlit-reveal

How to use

Here is an example app

import streamlit as st
import os
from cognite.streamlit import reveal
from cognite.client import CogniteClient

st.subheader("Cognite Reveal Streamlit example")
client = CogniteClient()
model_id = 123
revision_id = 234

selected_node_id = reveal(client, model_id, revision_id)
st.markdown("Selected node id: %d!" % int(selected_node_id))

Local development

It's recommended to add a clean environment. You need pip and node.

Clone repo git clone https://github.com/cognitedata/hackathon-reveal-streamlit.git

Install Python packages pip install streamlit

Install NPM packages and start server

cd cognite/streamlit/reveal/frontend
yarn
HTTPS=true yarn start

Then open https://localhost:3001/ to accept bad certificate.

Open repo folder in another terminal. Install this package as development package pip install -e .

Extract a token from Fusion, and start with

COGNITE_TOKEN="TOKEN" streamlit run examples/example.py

Local development in fusion stlite

Make sure you have set (reveal/init.py:8)[reveal/init.py:8] to _RELEASE = True.

Step 1) Build front end component with cd reveal/frontend && yarn && yarn build Step 2) Build streamlit component with python -m build (hint: pip install build) Step 3) Start local server python server.py

Open Fusion, create a Streamlit app and add the following the installed package http://localhost:8000/dist/reveal_streamlit_component-0.0.1-py3-none-any.whl

It will then load successfully inside Stlite.