Asking for 0 isotopes causes R crash

[wkumler]

Hi, awesome package! I've been using it for a while and just now tried to use it to identify MSMS fragment formulas, which means I don't need the isotope data. However, when I set maxisotopes=0, it causes R to crash.

E.g.:

library(Rdisop)
decomposeMass(mass = 147.0529, maxisotopes = 0)

It's not critically important because I can just ignore the isotope data, but took me while to debug and returning an error or warning instead would be super helpful.

[janlisec]

Hi @wkumler, I would implement a patch checking the maxisotopes parameter in R and return a warning, if specified as '0' instead of causing a crash.

However, I understood that your intention was to have the isotopes part of the function result to be omitted, correct? Why would you prefer this behavior?

[wkumler]

I mostly use Rdisop for mass decomposition, not isotope envelope prediction. My thought was that the algorithm could be sped up and the output cleaned up if the option to disable isotope envelope prediction was disabled. This is especially relevant for isotope-labeled experiments where the predictions aren't useful because I'm not working with natural abundances. So yes, while an error or warning thrown upon maxisotopes = 0 would be useful (and would have saved me an hour or two of frustrated debugging), a better solution would be to skip the envelope prediction part of the algorithm entirely when maxisotopes=0.

[janlisec]

I see, thanks for the clearification. I could implement the output clean-up in R (but you probably have done this as a post processing function already). I can not exclude the calculation from the C++ functions called underneath (as I am not fluent in C++). So excluding the isotope information would not speed up the function.

However, I actually use Rdisop because it is really fast already (compared to enviPat as the best alternative I am aware of). Still, a reasonable parameter setting should not lead to a crash and this will be fixed shortly.