Notice that it will take a few moments for Binder to install the Python environment. Also the application itself will be far less responsive then it would be if ran on a local computer. The delays are mainly caused by the network communications between the TTN prediction server (Google Cloud Run), the Python kernel (Binder), and the user's PC. Feel free to enable the verbose mode setting
verbose=True
to see how much time the application spends in the different stages. The prediction itself usually takes only a few milliseconds while request and rendering both take significantly longer (around two orders of magnitude).
conda create -n ttn-solid2d-ipynb python=3.6
conda activate ttn-solid2d-ipynb
conda install -c conda-forge jupyterlab
conda install -c conda-forge nodejs
conda install -c conda-forge matplotlib
conda install -c conda-forge ipywidgets
jupyter labextension install @jupyter-widgets/jupyterlab-manager
conda activate ttn-solid2d-ipynb
jupyter lab