Improve shadow price plotting module

[rouille]

Purpose

Improve plot_shadowprice_map module. Changes are mostly related to docstring

Testing

Ran associated notebook

Where to look

• plot_shadowprice_map:
  • format docstring
  • check type of parameters and raise errors
  • rename variable hour to date
  • change query for index in data frame and name of column
• shadowprice_map_demo:
  • import show from bokeh.io and encapsulate the call to plot_shadowprice
  • import pandas and use pd.Timestamp to pass the timestamp of the hour to plot to plot_shadowprice

Usage Example/Visuals

shadowprice_map_demo notebook

Time estimate

10 min

[BainanXia]

Looks good. Thanks.

[danielolsen]

There are a few places in the code where we use range(len(foo)), where I think enumerate would be cleaner.

[rouille]

There are a few places in the code where we use range(len(foo)), where I think enumerate would be cleaner.

I have simplified the for loops. In particular the one dealing with the color bar. The only difference with the previous version is that all bins spanning more than 5 $/MWh have size 5 now. Previously it would be only set for the first and last bins, resulting eventually in oversized bins in the middle (see plot below):

Before

After

[danielolsen]

It seems that this parameter is never actually used in this function, and the relevant info could be added to the bars dict within _construct_bus_legend, after the return from _get_bus_legend_bars_and_labels.

[rouille]

Yes I saw that. Same thing with bar_length_sum, it is equal to max(bars.values) and hence does not need to be returned.

[rouille]

Done

[danielolsen]

Clever.

[danielolsen]

Looks good, thanks.