Performance gains by avoiding the interpreter function

Question

I have this.

data = ResourceData["Epidemic Data for Novel Coronavirus COVID-19", 
   "USCounties"];

usaCountyCases[s_, c_] := 
 data[Select[MatchQ[Interpreter["USState"][s], #State] &]][
     Select[MatchQ[
        Interpreter["USCounty"][c <> " county, " <> s], #County] &]][
    All, #ConfirmedCases["LastValue"] &] // Normal // First

usaCountyCases["California", "Alpine"]

gives 124 (as of today). How do I speed this up? TIA.

Answer 1

1. You don't need to filter twice, by state and then by county, since the county already contains the state information.
2. The following modification to your function stuffs the string with the name of the county and the state into the appropriate Entity form, avoiding the calls to Interpreter.
3. The input strings are sanitized to remove spaces, so you can enter, for instance, "North Dakota", and the function will know to use "NorthDakota" in its query.
4. I've added the dataset as a parameter to your function so it is not hard-coded and you can use it on different datasets (e.g. subsets of the original).

ClearAll[usaCountyCases]
usaCountyCases[state_String, county_String, dataset_Dataset] := Module[{c},
  c = Entity[
        "AdministrativeDivision",
        {
          StringReplace[county, " " -> ""] <> "County",
          StringReplace[state, " " -> ""], "UnitedStates"
        }
      ];
  First[#, #]& @
   Normal@
    Query[
      Select[#County == c &], 
      #ConfirmedCases["LastValue"]&
    ][dataset]
]

usaCountyCases["California", "Alpine", data]
(* 124 *)

usaCountyCases["North Dakota", "Cass", data]
(* 54993 *)

From a performance perspective, these calls returned results practically instantaneously when I tried them.