Multiple patients have a height of -1

[JMathiszig-Lee]

Lots of patients, particularly the ones towars the end of the study have a height of -1. Plan email author of study to find out wtf

[tomdoel]

We'll need to modify the CSV reading code to either a) exclude patients with height -1, or b) do something sensible when a -1 appears (though I'm not sure what a sensible thing to do is). Presumably the heights must have been known at the time of the study!

[JMathiszig-Lee]

It's very frustrating as that dataset seems to be one of the best in terms of density of measurements so would be the most useful. I've just done a commit that modified the data set so that all unknown heights are set with the assumption of a BMI of 24.2 (average of the rest) so it now runs properly without errors. Obviously going forward this won't be valid and those patient will have to be removed when thinking about writing a paper!