arbitrary dimension names returned by griddap

[chrisdane]

Hi

I am confused how griddap() handles dimension names.

info <- rerddap::info("ncdcOisst21Agg_LonPM180")
varname <- "sst" # sea surface temperature
strsplit(info$alldata[[varname]]$value[1], ", ")[[1]] # dimension names returned by the `info` object
#[1] "time"      "zlev"      "latitude"  "longitude"
data <- rerddap::griddap(info, 
                         time=c("last", "last"), 
                         zlev=c("last", "last"), 
                         latitude=c("last", "last"), 
                         longitude=c("last", "last"), 
                         fields=varname)

The returned dimension names from the netcdf object are the same as from the info object:

names(data$summary$dim)
#[1] "time"      "zlev"      "latitude"  "longitude"

but the returned dimension names of the data.frame object storing the actual dimension and data values are different:

names(data$data)
#[1] "time" "lat"  "lon"  "zlev" "sst"

Suddenly "longitude" is now "lon" and "latitude" is "lat".

How am I supposed to use griddap in an automated way if suddenly names change?

Thanks a lot and kind regards,
Chris

Session Info

R version 4.1.0 (2021-05-18)
Platform: x86_64-conda-linux-gnu (64-bit)
Running under: Red Hat Enterprise Linux Server release 6.10 (Santiago)

Matrix products: default
BLAS/LAPACK: /sw/spack-rhel6/miniforge3-4.9.2-3-Linux-x86_64-pwdbqi/lib/libopenblasp-r0.3.12.so

locale:
 [1] LC_CTYPE=en_US.UTF-8       LC_NUMERIC=C              
 [3] LC_TIME=en_US.UTF-8        LC_COLLATE=en_US.UTF-8    
 [5] LC_MONETARY=en_US.UTF-8    LC_MESSAGES=en_US.UTF-8   
 [7] LC_PAPER=en_US.UTF-8       LC_NAME=C                 
 [9] LC_ADDRESS=C               LC_TELEPHONE=C            
[11] LC_MEASUREMENT=en_US.UTF-8 LC_IDENTIFICATION=C       

attached base packages:
[1] stats     graphics  grDevices datasets  utils     methods   base     

other attached packages:
 [1] microbenchmark_1.4.9 dplyr_1.0.7          rerddap_0.8.0       
 [4] oce_1.5-0            gsw_1.0-6            fields_13.3         
 [7] viridis_0.6.2        viridisLite_0.4.0    spam_2.8-0          
[10] ncdf4_1.19           colorout_1.2-2      

loaded via a namespace (and not attached):
 [1] Rcpp_1.0.8        pillar_1.6.4      compiler_4.1.0    tools_4.1.0      
 [5] digest_0.6.29     dotCall64_1.0-1   jsonlite_1.7.3    lifecycle_1.0.1  
 [9] tibble_3.1.6      gtable_0.3.0      pkgconfig_2.0.3   rlang_0.4.12     
[13] cli_3.1.1         DBI_1.1.1         curl_4.3.2        crul_1.1.0       
[17] gridExtra_2.3     stringr_1.4.0     xml2_1.3.3        rappdirs_0.3.3   
[21] generics_0.1.1    vctrs_0.3.8       maps_3.4.0        triebeard_0.3.0  
[25] grid_4.1.0        tidyselect_1.1.1  data.table_1.14.2 httpcode_0.3.0   
[29] glue_1.6.0        R6_2.5.1          fansi_1.0.2       hoardr_0.5.2     
[33] ggplot2_3.3.5     purrr_0.3.4       magrittr_2.0.1    urltools_1.7.3   
[37] scales_1.1.1      ellipsis_0.3.2    assertthat_0.2.1  colorspace_2.0-2 
[41] utf8_1.2.2        stringi_1.7.6     munsell_0.5.0     crayon_1.4.2

[rmendels]

@chrisdane Good question and I don't have an answer offhand. I am going to have to get into the code and see what is going on.

[rmendels]

@chrisdane - I am learning the inners of the code as I go, and lo and behold the "lon" and "lat" are baked into the code. I am not certain why this is so. So it is consistent in what it returns, but it is not consistent with the dataset info which is not good. I am pretty certain I know the fix, but it may take me a couple of days to implement it and test it.

Thanks for catching this.

[rmendels]

@chrisdane - Sorry this took so long. In wanting to fix this I wanted to allow 'rerddap' to "melt" other grids besides lat-lon. All of this is in 'rerddap v1.0.0' which is available from CRAN. The names in the returned structure now match those in other sources.