We use papermill
to run the template Jupyter notebook with parameters for a scripted generation of HF EDI tower data graphs for visual inspection.
Interesting side notes:
Pandas
read from MS SQL- several million datapoints, so Pandas plot with Matplotlib is slow
- so we convert to
Xarray
dataset, which adds more logic and efficiency