jupyter_notebook.md

How to use Pan3D and XArray Viewer in a Jupyter Notebook

Running Pan3D in a Jupyter notebook allows data scientists to incorporate the tool into their existing workflows and can facilitate greater collaboration between teammates. This tutorial assumes you have a running Jupyter notebook. You can find examples at notebook examples in the Pan3D code repository. You can run these examples on Binder here.

1. In your current kernel, install Pan3D:

   pip install pan3d[all]
   

2. In the first cell of your notebook, initialize Pan3D’s XArray Viewer.

   from pan3d.viewers.preview import XArrayViewer
   
   viewer = XArrayViewer()
   await viewer.ready
   

3. Prepare a configuration for the builder to import. This can come from a previously exported Pan3D configuration file. An example of this is shown in example_config_xarray.json:

   import json
   from pathlib import Path
   
   config = Path("../example_config_xarray.json")
   if config.exists():
       viewer.import_state(json.loads(config.read_text()))
       print("State loaded")
   else:
       print(f"Could not find example state in {config.resolve()}")
   

4. You can alternatively create a configuration dictionary. See url_config.ipynb for an example of this:

   config = {
       "data_origin": {
           "source": "url",
           "id": "https://ncsa.osn.xsede.org/Pangeo/pangeo-forge/noaa-coastwatch-geopolar-sst-feedstock/noaa-coastwatch-geopolar-sst.zarr"
       },
       "dataset_config": {
           "arrays": ["analysed_sst"],
           "slices": {
             "lon": [1000, 6000, 20],
             "lat": [500, 3000, 20]
           },
       },
   }
   
   viewer.import_state(config)
   

5. Finally to display the visualization in Jupyter, you can run the following cell

   viewer.ui
   

6. You can even share the visualization in PyVista as well by running the following cell assuming you've installed PyVista in your Python environment.

   import pyvista as pv
   
   plotter = pv.Plotter()
   plotter.show()
   
   xarray_reader = viewer.source
   actor = plotter.add_mesh(xarray_reader, scalars="analysed_sst", cmap='coolwarm')
   
   # sync viewer update
   viewer.ctrl.view_update.add(plotter.render)
   

While in the example we are using an XArrayViewer to drive our dataset builder, we can create such data source outside of the viewer and rely on it to drive any kind of visualization.