Combining hdf5 and csv VP files

[CeciliaNilsson709]

I downloaded hdf5 VP files from aloft to cover periods where csv VP files have not get been generated, but I am not able to combine them. With bind_into_vpts() I get the error:

Vertical profiles have different quantities

There seems to be several differences between the csv files and the hdf5 files, the csv have "data" quantities in different order, "vcp" had been added and "what" is missing from attributes.

Should we be able to combine VP files that are generated slightly differently?

library(dplyr)
#> 
#> Attaching package: 'dplyr'
#> The following objects are masked from 'package:stats':
#> 
#>     filter, lag
#> The following objects are masked from 'package:base':
#> 
#>     intersect, setdiff, setequal, union
library(bioRad)
#> Welcome to bioRad version 0.7.3
#> using vol2birdR version 1.0.1 (MistNet not installed)

tempdir <- tempdir()

hdf5_1 <- file.path(tempdir, "sehem_vp_20230505T235000Z_0xb.h5")
download.file(
  url = "https://aloftdata.s3-eu-west-1.amazonaws.com/baltrad/hdf5/sehem/2023/05/05/sehem_vp_20230505T235000Z_0xb.h5",
  destfile = hdf5_1
)

hdf5_2 <- file.path(tempdir, "sehem_vp_20230505T235500Z_0xb.h5")
download.file(
  url = "https://aloftdata.s3-eu-west-1.amazonaws.com/baltrad/hdf5/sehem/2023/05/05/sehem_vp_20230505T235500Z_0xb.h5",
  destfile = hdf5_2
)

vpts_hdf5<- bind_into_vpts(read_vpfiles(c(hdf5_1, hdf5_2)))

vpts_csv <-read_vpts("https://aloftdata.s3-eu-west-1.amazonaws.com/baltrad/daily/sehem/2023/sehem_vpts_20230506.csv")

bind_into_vpts(vpts_hdf5, vpts_csv)
#> Error in bind_into_vpts.vpts(vpts_hdf5, vpts_csv): Vertical profiles have different quantities

str(vpts_hdf5$data)
#> List of 15
#>  $ ff       : num [1:25, 1:2] NaN 7.75 15.34 NaN NaN ...
#>  $ dbz      : num [1:25, 1:2] -26.2 -16.9 -17.9 -26.5 -28.9 ...
#>  $ dens     : num [1:25, 1:2] 0.079 0.6637 0.5356 0.0727 0.0425 ...
#>  $ u        : num [1:25, 1:2] NaN -5.05 -10.6 NaN NaN ...
#>  $ v        : num [1:25, 1:2] NaN 5.88 11.09 NaN NaN ...
#>  $ gap      : num [1:25, 1:2] 1 0 0 1 1 1 1 1 1 1 ...
#>  $ w        : num [1:25, 1:2] NaN 105.2 31.2 NaN NaN ...
#>  $ n_dbz    : num [1:25, 1:2] 5160 29191 28710 25698 20488 ...
#>  $ dd       : num [1:25, 1:2] NaN 319 316 NaN NaN ...
#>  $ n        : num [1:25, 1:2] 20 364 196 66 50 7 14 5 0 0 ...
#>  $ DBZH     : num [1:25, 1:2] -26.18 5.19 -10.17 -18.62 -28.87 ...
#>  $ n_dbz_all: num [1:25, 1:2] 5160 30152 29100 25835 20488 ...
#>  $ eta      : num [1:25, 1:2] 0.869 7.301 5.892 0.8 0.468 ...
#>  $ sd_vvp   : num [1:25, 1:2] NaN 4.75 3.58 NaN NaN ...
#>  $ n_all    : num [1:25, 1:2] 20 371 200 71 50 7 14 5 0 0 ...

str(vpts_csv$data)
#> List of 16
#>  $ u        : num [1:25, 1:288] NaN -6.19 -10.36 -8.59 NaN ...
#>  $ v        : num [1:25, 1:288] NaN 3.37 11.35 10.9 NaN ...
#>  $ w        : num [1:25, 1:288] NaN 114.9 12.5 29.5 NaN ...
#>  $ ff       : num [1:25, 1:288] NaN 7.05 15.36 13.87 NaN ...
#>  $ dd       : num [1:25, 1:288] NaN 299 318 322 NaN ...
#>  $ sd_vvp   : num [1:25, 1:288] NaN 4.55 3.15 3.9 NaN ...
#>  $ gap      : num [1:25, 1:288] 1 0 0 0 1 1 1 1 1 1 ...
#>  $ eta      : num [1:25, 1:288] 4.8 7.59 2.94 1 1.18 ...
#>  $ dens     : num [1:25, 1:288] 0.4364 0.6902 0.2675 0.0911 0.1077 ...
#>  $ dbz      : num [1:25, 1:288] -18.8 -16.8 -20.9 -25.6 -24.8 ...
#>  $ n        : num [1:25, 1:288] 30 330 173 68 46 4 6 1 2 0 ...
#>  $ DBZH     : num [1:25, 1:288] -18.76 2.82 -11.66 -16.97 -24.83 ...
#>  $ n_dbz    : num [1:25, 1:288] 5129 29912 28963 25635 20479 ...
#>  $ n_all    : num [1:25, 1:288] 30 336 175 70 46 4 6 1 2 0 ...
#>  $ n_dbz_all: num [1:25, 1:288] 5129 30125 29112 25834 20479 ...
#>  $ vcp      : num [1:25, 1:288] NA NA NA NA NA NA NA NA NA NA ...

^{Created on 2024-02-09 with reprex v2.1.0}

[iskandari]

@CeciliaNilsson709 good observation. Ideally they should be combinable, even if some optional fields are missing in one vpts object and not in the other