[Calphad 2021] Update pycalphad agenda

[bocklund]

No description provided.

[bocklund]

I'd like to see a little more time dedicated to getting people comfortable with the returned xarray Datasets, if possible. Inspecting them, but also using them for plotting and common operations. Some places where 2020 workshop's 03 Calculation Results.ipynb can be improved:

1. it can be difficult to understand the multi-dimensionality, dimensions, coordinates, data variables. We should try to make this clear in fewer words. Notebook 03 is a little too wordy for a workshop tutorial and is more reference material (my fault - I wrote it). Maybe borrow from xarray's images or tutorials? Could use a link to the xarray documentation too.
2. explanation of how pycalphad pads the arrays with nan or '', how to tell the difference from convergence failures
3. avoid dropping down into NumPy arrays and preserve the xarray semantics (see also: DOC: regenerate examples and small fixes pycalphad/pycalphad#321)
4. some worked examples of how to inspect the datasets/dataarray values
5. worked examples that specifically illustrate how to plot data and issues around dataarray shapes (use of squeeze, broadcast_like)
6. some discussion on how where works and worked examples for filtering for certain criteria (see cookbook examples) - make the connection to "screening" for phases, properties (NP), etc. of interest
7. How to save the data: CSV and/or NetCDF