[R] Refer to the Arrow C++ compute engine as Acero

[ianmcook]

Once ARROW-16632 and ARROW-16626 are complete, we should update some of the text in the R cookbook to refer to the Arrow C++ compute engine using its new name, Acero. We should try to be quite explicit about the fact that the {arrow} R package converts {dplyr} code into an Acero ExecPlan.

[thisisnic]

Happy to update "Arrow C++ compute engine" to "Acero" but can you expand upon where

We should try to be quite explicit about the fact that the {arrow} R package converts {dplyr} code into an Acero ExecPlan.

would fit, in terms of where in the cookbook it'd be relevant to mention this? I'm not entirely convinced that this is the right place to be talking about ExecPlans as it feels like a fairly technical level of detail that belongs in other documentation (though open to being persuaded otherwise!)

[ianmcook]

can you expand upon where

We should try to be quite explicit about the fact that the {arrow} R package converts {dplyr} code into an Acero ExecPlan.

would fit, in terms of where in the cookbook it'd be relevant to mention this?

Yes, I think it would be valuable to add some content about this to the Manipulating Data - Tables R cookbook entry. It would be great if that cookbook entry could explain what happens under the hood when the user calls dplyr verbs on an Arrow table/recordbatch/dataset and then calls collect() or compute(): an Acero ExecPlan is constructed and executed.

[thisisnic]

OK, makes sense, we could mention that briefly in the intro and then link to some docs where it's explained in more detail.