And so we can see that this general trend of sales over time

is being followed by the three different products, but it's

more accentuated with one of the products than it is perhaps with the other two.

And so

that makes sense to concatenate a time axis with a product axis in this case.

It may not make sense to concatenate a quarter axis with

a month axis because they're both time dimensions, and

so you'd be seeing the same plot here over the demarcations of quarter and

then the same plot over the demarcations of a month.

And you'd be using half of the space to build this up, and

so it doesn't make as much sense in that case.

We can also take two dimensions and

create an axis from the product of those two dimensions.

For example, if we have the data over time, over four quarters.

And we have the data over products instead of showing them side-by-side,

we might show the data for the coffee, espresso and tea, the product values for

quarter one, quarter two, quarter three, quarter four.

And so this product basically gives you every combination of coffee and

quarter so that you can see how your data's performing over time and

product simultaneously, with a single axis here representing these two

dimensions by looking at all the combinations.

Similarly, we could look at quarter times month.

This doesn't make much sense either, because we'd be looking at the first

quarter, which should represent just the first three months and yet

I'm plotting all 12 months here.

And so just doing a simple product of two time axes doesn't really

help, because I'd only have data for the first quarter showing up in

the first three months, and then the other nine months would basically be

zero valued because there wouldn't be data for these months.

That was also in the first quarter.