[1m[36m> Draw me a plot so I can see what this plate looks like.[0m

Two things have to exist before a picture can: a plot card draws a block's output,
so there has to be a block, and it has to have run.

[32m* write_workflow(path='workflows/main.yaml', ...)[0m
[2m    load-plate  (Load)  ->  data/raw/drug_A-J_1mM_500uM_200uM_100uM_50uM_20...csv[0m

One block for now. Everything else gets added to this same workflow later.

[32m* list_plot_targets(workflow_path='workflows/main.yaml')[0m
[2m    main:load-plate:data    Load -> data    (DataFrame)[0m
[32m* scaffold_plot(plot_id='plate_raw', target_id='main:load-plate:data')[0m
[2m    plots/plate_raw/plot.yaml[0m
[2m    plots/plate_raw/render.py[0m

One square per well, one column per group, coloured by viability. Ninety-six
numbers in a column tell you very little; ninety-six squares laid out the way the
plate was laid out tell you a lot, because a bad well is something that happened in
a place.

The x axis is the group label, because that is every bit of what the table knows
about which well got what.

[32m* run_workflow(path='workflows/main.yaml')[0m
[2m    run queued[0m

Running it, because the card needs that block's output and the block has to have
produced one. Open the Plots tab and press Run on the card.
