Refining GNsearch results

You should notice immediately that along with latitude and longitude, the results include a lot of information about the various places under some unclear column names. We can use this information to narrow our search results. An overview of each of the parameters – column names – is available here.

Not always do column/parameter names match up with arguments to the GNsearch function. For example, featureClass is the argument that maps to the fcl parameter. For many uses this is probably the finest search parameter one might use: P = populated place, L = parks, etc. Much much finer designations can be given with the fcode or fcodeName parameter (see here ). For example, Fargo, ND and Fargo, AR are both P for populated places, but only Fargo, ND is fcode PPLA2, for “seat of a second-order administrative division”, or county seat. (States are first-order administrative divisions).

Because we know we want the Fargo in the state of North Dakota, we specify adminCode1=“ND” and only one result is returned:

GNsearch(name_equals = "Fargo", adminCode1="ND")

Cut to the chase

Here’s the script used to generate the list of towns and RECs we used in class nd.features.d:

nd.features <- data.frame(feature=c("city", "city", "city", "city", 
                                    "rec", "rec", "rec", "rec", 
                                    "rec", "rec"), 
                          name=c("Fargo", "Grand Forks", "Bismarck", "Minot", 
                                 "Hettinger", "Williston", "Carrington","Dickinson",
                                 "Streeter", "Langdon") )
nd.features.d <- 
  nd.features %>%
    split(.$name) %>%
      map( ~ GNsearch(name_equals = .x$name, country = "US", 
                      adminCode1="ND", featureClass="P")) %>%
        map_dfr(~ (.))  %>%
    full_join(nd.features) %>%
    mutate(state = adminName1, 
           long=as.numeric(lng), 
           lat = as.numeric(lat)) %>%
    select(feature, name, state, long, lat) 

NDs <- map_data("state") %>%
  filter(region == "north dakota")

NDc <- map_data("county") %>%
  filter(region == "north dakota")

ggplot() + theme_bw() + 
  geom_path(data=NDc, 
            aes(x=long, y=lat, group=group), 
            color="dark grey") +
  geom_path(data=NDs, 
            aes(x=long, y=lat, group=group), 
            color="black") +
  geom_point(data=nd.features.d, 
             aes(x=long, y=lat, 
                shape=feature, color=feature)) +
  geom_text(data=nd.features.d, 
            aes(x=long, y=lat, 
                label=name), nudge_y = 0.15) +
  coord_map("conic", 60)