Skip to content

Sub-seasonal daily data: aggregation, saving and visualizing

Hi @vagudets,

When using the Crossval_calibration() workflow, with the recipe Probabiltiessave: 'all'

This piece of code fails:

 info(recipe$Run$logger,
           "SAVING OBSERVED CATEGORY LIMITS")
    save_percentiles(recipe = recipe, percentiles = tmp_lims2,
                       data_cube = data$obs,
                       agg = "global", outdir = NULL)
Error in ArrayToNc(vars, outfile) : 
  The dimension 'time' is defined or used more than once in the provided data but the dimension specifications do not match.

In my case, I am using sub-seasonal daily data and applying the time Aggregation() from 1 to 7.

The save_percentiles() function calls the times <- .get_times(recipe, data_cube, fcst.sdate, calendar, init_date) returns an object of length 7 when it should be of length 1, I guess:

$time
[1] 7
attr(,"variables")
attr(,"variables")$time
attr(,"variables")$time$units
[1] "days since 2024-10-24T00:00:00"

attr(,"variables")$time$calendar
[1] "proleptic_gregorian"



$time_bnds
     [,1] [,2] [,3] [,4] [,5] [,6] [,7] [,8] [,9] [,10] [,11] [,12] [,13] [,14]
[1,] -310 -238 -142 -238 -142  -70 -142  -70   26   -70    26    98    26    98
[2,] -166  -94    2  -94    2   74    2   74  170    74   170   241   170   241
     [,15] [,16] [,17] [,18] [,19] [,20] [,21] [,22] [,23] [,24] [,25] [,26]
[1,]   193    98   193   265   193   265   361   265   361   433   361   433
[2,]   337   241   337   409   337   409   505   409   505   577   505   577
     [,27]
[1,]   529
[2,]   673
attr(,"variables")
attr(,"variables")$time_bnds
attr(,"variables")$time_bnds$units
[1] "days since 2024-10-24T00:00:00"

attr(,"variables")$time_bnds$calendar
[1] "proleptic_gregorian"

attr(,"variables")$time_bnds$long_name
[1] "time bounds"

Probably, this case has not been considered yet. I let this issue here in case we could apply some tests for Aggregation() and saving modules.

Thanks,

Núria