Aggregation & missing values

[aukkola]

When aggregating to a coarser time step, the code sets the new time step to NA if any of the original time steps are NA (controlled by na.rm=FALSE):

aggr_data <- sapply(seq, function(x) mean(aggr_data[x:(x+ntsteps-1)], na.rm=FALSE))

This means that when aggregating from halfhourly to daily, for example, only 1 out of 48 time steps needs to be missing for the whole day to be set to missing. This seems overly strict, with no flexibility for the user.

Suggest adding a new processing threshold for controlling the percentage of aggregated time steps allowed to be missing.