Preparing data for caffeine upload
Math: everything looks perfect
another comment/ understanding issue:
For very small values of concentrations, is it not possible that the uncertainty value underflows, thus goes to 0 and it would make more sense of handling them in percentages? maybe that could also pose a problem for simulation?
Other than that I cannot find any problems.
Thanks for the feedback, will change the water content to that reference as it looks more reliable. Table 2 says it's 70%, so I'll just use that.
Regarding your other comment, do you mean cases where the uncertainty value is above the measurement itself, i.e. CV > 100%? I have kept all of those cases for now, but assessing the effect of filtering those values out could be something to look further into in the future.