Closed jondea closed 6 years ago
How about WaveMoments for the name of the type? That suggests it has moments + metadata
But we haven't used the word wave anywhere else, it would be odd to only use it here in relation to FrequencyModels and TimeModels
I'm wondering if we shouldn't rename those too. I mean "Model" is very vague. What if instead we used FrequencyMST, TimeMST and MomentsMST ? With MST meaning Multiple ScaTtering? That way we only have one keyword MST throughout our package, and it avoids any future conflict with other packages. I mean the word TimeModel could be used in so many other circumstances.
I agree, although I'm not sure I like MST, especially given that our package is called MultipleScattering so their full names would be MultipleScattering.FrequencyMST and MultipleScattering.TimeMST. I wouldn't worry about name conflicts, because if they exist then people can use the full names. I think clarity and meaning is more important.
Then how about replacing MST with just Wave, or Scattering? So FrequencyWave, TimeWave, MomentsWave?
We decided on the names for the types: FrequencySimulation and TimeSimulation, as it makes it clear that these are objects with all the information necessary to do a whole simulation. For the moments type we choose StatisticalMoments to clearly differentiate it from other definitions of moments.
It's odd that Moment has a subtype which is called moments. It would make more sense to call it label+moments or moments with metadata. Another idea would be to borrow the word from Summary Statistics and call it a ModelSummary.