Multiple rcs values when downloading RMI data

[bart1]

getRad::get_vpts("behel", c("2025-03-18","2025-03-21"), source = "rmi")
#> Warning in check_multivalue_attributes(data): multiple rcs values found,
#> storing only first (11.0133105629406) as the functional attribute.
#>                    Irregular time series of vertical profiles (class vpts)
#> 
#>            radar:  behel 
#>       # profiles:  1108 
#> time range (UTC):  2025-03-18 00:00:00 - 2025-03-21 23:55:00 
#>    time step (s):  min: 300     max:  4500

Where we get a warning from bioRad that the RCS values are varying. I think this because for each day the RCS is calculated separately here:

getRad/R/utils_vpts_rmi.R

Line 128 in f4bc216

rcs = calc_single_mean_rcs(.data$eta, .data$dens)

It would be good to prevent this I think as it is most likely only a result of rounding errors. We probably should check that the variation is in deed minute before averaging to one value. @PietrH could you pick this up? No hurry

[PietrH]

So, instead of calculating it for every record. You want one value that is the average (mean).

• Except, if there is significant variation. How do you want to quantify this? Is using the median enough?
• If there is significant variance, what should we do? Return a warning? Drop the record with a warning?

[bart1]

So, instead of calculating it for every record. You want one value that is the average (mean).

I is already average now, just per day. So you don't get the warning from bioRad if you request only one day.

* Except, if there is significant variation. **How do you want to quantify this?** Is using the median enough?

I would explore a bit around, my guess is a tolerance of .1% or 1% is enough to avoid this warning for cases where the differences purely stem from rounding errors. I think it is unlikely RMI deliberately varies RCS ;).

* If there is significant variance, what should we do? Return a warning? Drop the record with a warning?

I would say warn is good enough