improvement request: remap days 0-6 to human-readable labels (0 and 6 is the weekend indeed, americans like to put part of it at the weekstart) the problem with this chart is a lot of overplotting (google the images on the term if unclear), there is a relatively easy way around it. search for "density contours" when you create a new cell here. see how it's made with a built-in example. then convert your chart cell to JS, there is a button underneath when the cell is in focus. you only need to replace one word in the resulting code to make it work as a density contour chart, see from the example you searched the nominal color palette would perhaps need to be more distinct for the contours, red and orange are a bit too close, you can also experiment with "fill" and "opacity" properties

i think "total" is in dollars though, so you are looking at revenue. if you want orders you would need to switch Y enc also you need to add aggregation, such as "sum" and switch the mark type

good that you bring all datapoints into one picture! an overplotting here can be avoided with a stacked area chart: switch mark type to area, set Y to SUM of orders. better? lemme see you try further suggestion: let's make a similar analysis but separately by pizza name and size. currently them two are in a single value, can you instead add 2 extra columns name and size? then facet both of them and suddenly the data makes much more sense! when you are stuck you can see how i done it in my account with a similar assignment