re-factor mw_wrangle_wfhz() account for children measured lying down/recumbent length

[tomaszaba]

Sometimes recumbent height is taken on some children. A correction factor is applied in those cases. Need to refactor mw_wrangle_wfhz() function to that end.

[tomaszaba]

@ernestguevarra, on this issue, in ENA, one would only indicate "l" for length, to mean recumbent, for a child aged 24-59 who for some reason was not measured standing, as it should. This means that all other children whose expectation was not violated, are left blank. I was reading the {zscorer} and I could not grasp clearly how the standing argument work. How should the vector look like?

[ernestguevarra]

@tomaszaba, the standing variable (if there is) should be coded as 1 for standing, 2 for supine and 3 for unknown. If no standing variable is provided, then all values for standing are set to unknown. if unknown, then the age rule comes into play and age will determine whether the measurement is a standing or supine.

[ernestguevarra]

this is the same algorithm used by the anthro package by WHO (they copied this from us)

[ernestguevarra]

you don't need to do anything different when you use zscorer. If you have data on whether standing or not, then use that but coded as above and it will take care of unknown. If there is no standing variable which is the default, then zscorer will deal with that.

[tomaszaba]

OK. Thanks. Let me find an ENA like dataset to show how they deal with this issue and then we see how we can address that in this issue. Will come back to this tomorrow.

[ernestguevarra]

@tomaszaba, this is another package that we need to benchmark SMART ENA and zscorer against in terms of z-score calculations - https://github.com/ropensci/gigs