Support interval+reduce in the area mark

[Fil]

Defaults for x, y, indexOf, and identity are now set inside maybeDenseIntervalX / maybeDenseIntervalY, instead of being scattered across the mark constructors. This fixes a bug where areaY (and areaX) with a dense interval did not apply the reducer to the y (or x) channel, because the default was set too late — after the dense interval transform had already been configured.

For example, this now works as expected:

Plot.areaY(data, {x: "date", interval: "day", reduce: "count"})

Previously, the dense interval would not reduce y because y was still undefined when maybeDenseIntervalX ran. The identity default was only applied later by maybeIdentityY.

Closes #2328

[mbostock]

More description about how this fixes the problem would be helpful in the PR. (And it would be more relevant than your experience with AI, though I appreciate you sharing!) I guess this changes the behavior exactly when y is undefined, but preserves the current behavior when y is null. Will anything else be affected by this fix?

[Fil]

Good call. I added details in the description.

[mbostock]

I figured out how to lift the defaults for dense intervals up so they are shared between line and area. That feels a little cleaner to me.

[mbostock]

I’m less sure about this, but at least according to the tests, the area mark also doesn’t need the maybeIdentity transform now?

[Fil]

Merging main was a bit complicated by 3542325; then I saw that lineX could be simplified to follow the same pattern as we had in areaX (we were setting the defaults x = indexOf, y = identity in two places). PTAL?