Multiple stationids in ncdc_stations

[tchakravarty]

I am unable to figure out how to specify a list of weather stations in the ncdc_stations function. The function has a relevant sounding argument stationid, but it asks to pass a chain of of [sic] station ids in a comma-separated vector. I don't understand what data structure is expected here.

Here is what I have in my code:

# get country data
df_ghcnd_countries = 
  read_table("http://www1.ncdc.noaa.gov/pub/data/ghcn/daily/ghcnd-countries.txt", 
             col_names = c("Abbreviation", "Country"))

# get station data
df_ghcnd_stations = 
  read_fwf("http://www1.ncdc.noaa.gov/pub/data/ghcn/daily/ghcnd-stations.txt", 
           col_positions = fwf_positions(
             start = c(1, 13, 22, 32, 39, 42, 73, 77, 81),
             end = c(11, 20, 30, 37, 40, 71, 75, 79, 85),
             col_names = c("ID", 
                           "Latitude",
                           "Longitude", 
                           "Elevation",
                           "State",
                           "Name",
                           "GSN Flag",
                           "HCN/CRN Flag",
                           "WMO ID")
           )
  )

# get the list of stations for Germany; using a positive lookahead regex
german_stations = df_ghcnd_stations$ID[grepl(paste0("^", df_ghcnd_countries$Abbreviation[
  df_ghcnd_countries$Country == "Germany"
], "(?=[0-9])"), df_ghcnd_stations$ID, perl = TRUE)]

# get the weather station data
station_list = ncdc_stations(
  startdate = "2014-01-01",
  enddate = "2015-11-01",
  stationid = paste0(paste0("GHCND:", german_stations), collapse = ";")
) 

This returns with just data for one (the first) station.

I have also tried paste0(paste0("GHCND:", german_stations), collapse = ",") and debugging the function, but since I do not know the URL formation requirements for NOAA, I am unable to proceed further. Any help here would be appreciated.

Thanks.

[sckott]

Thanks for your question @tchakravarty I'll have a look soon

[sckott]

• We do have a function in rnoaa that does what your df_ghcnd_countries variable gets, see ghcnd_countries()
• We do have a function in rnoaa that does what your df_ghcnd_stations variable gets, see ghcnd_stations()
• I mistakingly allowed more than one station to be passed to ncdc_stations() - will push a change soon to fix that, it now stops with error if stationid length is > 1 - You can instead use e.g., lapply() or a for loop to deal with many station ids

Try this at the end of your script above instead of the call to ncdc_stations()

dplyr::bind_rows(lapply(paste0("GHCND:", german_stations), function(z) {
  ncdc_stations(
    startdate = "2014-01-01",
    enddate = "2015-11-01",
    stationid = z)
}$data))
#> Source: local data frame [22 x 9]
#> 
#>    elevation    mindate    maxdate latitude                  name datacoverage                id
#>        (int)      (chr)      (chr)    (dbl)                 (chr)        (dbl)             (chr)
#> 1         62 1891-01-01 1991-12-31  51.9506          MUENSTER, GM       0.9802 GHCND:GM000001153
#> 2          4 1890-01-01 2016-02-09  53.0464            BREMEN, GM       0.9928 GHCND:GM000001474
#> 3        144 1907-05-01 2003-05-19  49.7517             TRIER, GM       0.8674 GHCND:GM000002277
#> 4        285 1901-01-01 2015-11-30  49.4253    KAISERSLAUTERN, GM       0.8557 GHCND:GM000002288
#> 5        112 1876-01-01 2008-10-31  49.0392         KARLSRUHE, GM       0.9931 GHCND:GM000002698
#> 6        401 1900-01-01 2016-02-09  48.7703         STUTTGART, GM       0.9663 GHCND:GM000002716
#> 7         59 1890-01-01 2015-11-30  53.6442          SCHWERIN, GM       0.9815 GHCND:GM000003038
#> 8         93 1900-01-01 2015-03-22  51.5144             HALLE, GM       0.9928 GHCND:GM000003218
#> 9        246 1917-01-01 1999-12-31  51.1167 DRESDEN WAHNSDORF, GM       1.0000 GHCND:GM000003244
#> 10        51 1876-01-01 2014-06-02  52.4639     BERLIN DAHLEM, GM       0.9748 GHCND:GM000003319
#> ..       ...        ...        ...      ...                   ...          ...               ...
#> Variables not shown: elevationUnit (chr), longitude (dbl)

Note that there is a problem however with some other ncdc_*() functions (ncdc_stations only accepts one at a time) in that they are supposed to take in many stationid parameters, but they don't. I've contacted NOAA to see if they can fix.

[sckott]

closing, see #150 for following up with NOAA about the multiple stationids issue