a challenge for higher points: let's add nicer labels for days of week

another challenge: do this but split names to categories and sizes, then use facet encodings (fx,fy) to arrange the charts

another challenge: let's do analysis of individual datapoints, like orders over time by pitsa type by size by state etc. show the original unaggregated data in a way that it makes sense

it is great that you've learned d3.rollup — a versatile tool to do all sorts of data transformations, so your method was absolutely correct. i just wanted to show you an alternative solution here, that uses Plot built-in data transformation feature. I usually forget the syntax, like what should be X and Y in Plot.groupX({ y:, so i use a chart cell to give me a template of this or look it up in the docs