Output of accessor function as NumericList ?

[philouail]

If we follow the Spectra package structure the accessor function rtime() and intensity() should give an ouput as a NumericList

I always will check in the validPeaksData() function that we have numeric data anyway so is it necessary ? NumericList are a bit less efficient.

just wondering if there is any reason to keep it that way and what would be the drawbacks to just have a list ?
Could a future ChromBackend implementation need this ?

[jorainer]

I have no strong opinion here - with a slight preference to use list of numeric instead of NumericList.

[philouail]

Some extra info: here are the results of a microbenchmark with and without having a NumericList as an output instead of normal list output.... It's not just a bit less efficient, this is pretty consequential.

[jorainer]

I would go for a list of numeric for now. The backend needs to ensure that the list it returns is indeed a list of numeric (not some random logical or character in it), but that's something that the backend can check e.g. when initialized.

Worst case we can always wrap an additional NumericList() call around the returned list...

it is just important that we properly define and document the (expected) type of the return value of methods (this should go into the documentation of the ChromBackend.

[philouail]

Yes as of now I find it more easy to implement and it seems that it is just more efficient in general. Also I implement a .CORE_PEAKS_VARIABLES which defines the names and class of these, which makes it easy for me to check that they are indeed numeric.

If something comes up I'll add to the discussion :)

[jorainer]

Excellent, yes, I agree to split/distinguish between core chromatogram variables and core peaks variables.