[Bug][Jupyter Widget] Error displaying widget: model not found

Describe the bug Attempting to display a map in JupyterLab 2.1.0 prints an error within Jupyter Lab: “Error displaying widget: model not found”

My guess (but I really am guessing!) is that the keplergl widget needs to be updated to play nicely with JupyterLab 2.1.0. (Or maybe I’m doing something stupid! This is the first time I’ve tried installing the keplergl widget!)

To Reproduce

Installation:

conda config --add channels conda-forge
conda create --name test_kepler python=3.8 jupyterlab ipywidgets nodejs
conda activate test_kepler
pip install keplergl
jupyter labextension install @jupyter-widgets/jupyterlab-manager keplergl-jupyter
jupyter lab

Create a new notebook and run:

from keplergl import KeplerGl
map_1 = KeplerGl()
map_1

Output:

User Guide: https://github.com/keplergl/kepler.gl/blob/master/docs/keplergl-jupyter/user-guide.md
Error displaying widget: model not found

Expected behavior A map should be displayed 😃

Environment:

• conda virtual environment
• Python version: 3.8.2
• keplergl Widget version 0.1.2
• jupyterlab 2.1.0
• ipywidgets 7.5.1

jupyter labextension list

JupyterLab v2.1.0
Known labextensions:
   app dir: /home/jack/miniconda3/share/jupyter/lab
        @jupyter-widgets/jupyterlab-manager v2.0.0  enabled  OK
        jupyterlab-plotly v4.6.0  enabled  OK
        keplergl-jupyter v0.1.2  enabled  OK

Bug reports in other projects Here are issues in other projects where users report seeing the “Error displaying widget: model not found” issue:

• https://github.com/matplotlib/ipympl/issues/112
• https://github.com/quantopian/qgrid/issues/286

[heshan0131]

just published keplergl==0.2.0 with support for JupyterLab 2.0

[ramayer]

If anyone’s stuck on an old version of the components, this workaround worked for me:

test_map = keplergl.KeplerGl(height=600)
orig_html = str(test_map._repr_html_(),'utf-8')

better_html = orig_html
import base64
b64 = base64.b64encode(better_html.encode('utf-8'))
src = f"data:text/html;base64,{b64.decode('utf-8')}"
base64d_html = f'<iframe src="{src}" style="width:95%; height: 600px">'
import IPython
IPython.display.HTML(base64d_html)