display of plotting output in tutorial notebooks is inconsistent

[Thomas Gilgenast]

in the tutorial notebooks, we sometimes want to demonstrate the use of the outfile kwarg to write plot output to a file (showing it later using IPython.display.Image), while at other times we want to show the plot inline with %matplotlib inline

we currently use %matplotlib notebook to try to hide the plots that we are writing to file from being displayed inline in the notebook

this works on colab, but in a normal jupyter environment it shows the large interactive plotting UI (one per rep if parallelized) with no content (since the content is in a different figure that has been saved to disk by now)

we consider three possible solutions to hiding the plot output for these cells:

1. %matplotlib agg: this does exactly what we want but it was only added in IPython 6.0 which only supports Python 3, see this issue
2. plt.close(): this hides the UI but leaves behind a giant rectangle that is tricky to collapse programmatically, see this issue
3. %%capture: this feels like an indirect way of solving the problem but it works on colab and in jupyter, see this SO answer

[Thomas Gilgenast]

wip on notebook building

closes #57

added gmean_20_8 binning to default pipeline config so we can use it in the tutorial

temporarily changed tutorials/run.py to run only the scripting tutorial for testing purposes

first stab at tutorial build pipeline in bibucket

→ <<cset 93e277d38bf8>>

[Thomas Gilgenast]

we ended up going with the %%capture strategy in the commit referenced above