gamlss has trouble doing predictions without the original training data.
Usage
predict_centiles(newdata, model, centiles = c(5, 10, 50, 90, 95), ...)
pivot_centiles_longer(data)
Arguments
- newdata
a one-column dataframe for predictions
- model
a gamlss model prepared by
mem_gamlss()
- centiles
centiles to use for prediction. Defaults to
c(5, 10, 50, 90, 95)
.- ...
arguments passed to
gamlss::centiles.pred()
- data
centile predictions to reshape for
pivot_centiles_longer()