Combine vaccine suppquals with parent domains for consistency with all the other admiral packages

[rossfarrugia]

All the suppxx_vaccine suppqual datasets should be combined with their parent domains, in line with this update to admiralvaccine:
pharmaverse/admiralvaccine#246

FYI @ahasoplakus @manciniedoardo

[bundfussr]

@rossfarrugia , oh, is this a general rule?

We have also suppqual datasets for onco but they are not combined here. If combining them is necessary, it is done in the vignette using metatools. Therefore metatools is not a dependency but listed in Suggests: only.

There are more suppqual datasets available, e.g., suppae, suppdm, ... It seems that they are separate and not combined.

Or did I misunderstood the issue?

[rossfarrugia]

Ah OK am I mistaken then here @bms63 & @manciniedoardo? I thought we only ever used {metatools} in our E2E Examples (https://github.com/pharmaverse/examples) and we kept this purposefully separate from {admiral} packages.

@ahasoplakus did actually offer the option of moving {metatools} to Suggests: only in {admiralvaccine} so maybe thats a less painful fix.

However note that there are breaking changes for combine_supp() in latest {metatools} release, so we still need to likely check and make updates for this if we keep it.

[bms63]

oh that is unfortunate on the breaking changes for combine_supp().

i like the admiralonco method - sounds like admiralvaccines should just adopt that?

[rossfarrugia]

Fine for me if all agree!

[bundfussr]

oh that is unfortunate on the breaking changes for combine_supp().

The breaking change is:

Breaking change: combine_supp() requires that the QNAM columns are not in the source dataset

This seems a quite unusual scenario to me. Thus I would not expect that we need to update on our side.

[rossfarrugia]

@bundfussr as FYI the reason this first came to my attention is that @millerg23 mentioned that {metatools} latest CRAN submission had failed due to it breaking {admiralvaccine}

[bundfussr]

@bundfussr as FYI the reason this first came to my attention is that @millerg23 mentioned that {metatools} latest CRAN submission had failed due to it breaking {admiralvaccine}

Hm, if this is the reason for breaking {admiralvaccine} I wonder if the vaccine SDTMs are correct. I think the supp dataset should only provide variables which are not already in the parent domain.