GeoDB: Prepare for population

src/instance/geodb.jl

@@ -17,7 +17,13 @@ struct GeoPoint
     lon::Float64
 end
 
-DB_CACHE = Dict{String,Dict{String,GeoPoint}}()
+struct GeoRegion
+    centroid::GeoPoint
+    population::Int
+    GeoRegion(; centroid, population) = new(centroid, population)
+end
+
+DB_CACHE = Dict{String,Dict{String,GeoRegion}}()
 
 function centroid(geom::Shapefile.Polygon)::GeoPoint
     x_max, x_min, y_max, y_min = -Inf, Inf, -Inf, Inf
@@ -32,7 +38,7 @@ function centroid(geom::Shapefile.Polygon)::GeoPoint
     return GeoPoint(round(y_center, digits = 5), round(x_center, digits = 5))
 end
 
-function download_census_gov(url, outputdir, expected_crc32)::Nothing
+function download_zip(url, outputdir, shp_crc32)::Nothing
     if isdir(outputdir)
         return
     end
@@ -40,7 +46,7 @@ function download_census_gov(url, outputdir, expected_crc32)::Nothing
     @info "Downloading: $url"
     zip_filename = download(url)
     actual_crc32 = open(crc32, zip_filename)
-    expected_crc32 == actual_crc32 || error("CRC32 mismatch")
+    shp_crc32 == actual_crc32 || error("CRC32 mismatch")
     open(zip_filename) do zip_file
         zr = ZipFile.Reader(zip_file)
         for file in zr.files
@@ -52,17 +58,17 @@ function download_census_gov(url, outputdir, expected_crc32)::Nothing
     return
 end
 
-function load_census_gov(;
+function geodb_load_gov_census(;
     db_name,
-    url,
-    expected_crc32,
-    shp_filename,
     extract_id,
-)::Dict{String,GeoPoint}
+    shp_crc32,
+    shp_filename,
+    shp_url,
+)::Dict{String,GeoRegion}
     basedir = joinpath(dirname(@__FILE__), "..", "..", "data", db_name)
     csv_filename = "$basedir/locations.csv"
     if !isfile(csv_filename)
-        download_census_gov(url, basedir, expected_crc32)
+        download_zip(shp_url, basedir, shp_crc32)
         @info "Processing: $shp_filename"
         table = Shapefile.Table(joinpath(basedir, shp_filename))
         geoms = Shapefile.shapes(table)
@@ -77,8 +83,12 @@ function load_census_gov(;
     end
     if db_name ∉ keys(DB_CACHE)
         csv = CSV.File(csv_filename; types = [String, Float64, Float64])
-        DB_CACHE[db_name] =
+        DB_CACHE[db_name] = Dict(
-            Dict(row.id => GeoPoint(row.latitude, row.longitude) for row in csv)
+            row.id => GeoRegion(
+                centroid = GeoPoint(row.latitude, row.longitude),
+                population = 0,
+            ) for row in csv
+        )
     end
     return DB_CACHE[db_name]
 end
@@ -87,13 +97,13 @@ function _id_2018_us_county(table::Shapefile.Table, i::Int)::String
     return table.STATEFP[i] * table.COUNTYFP[i]
 end
 
-function load_2018_us_county()::Dict{String,GeoPoint}
+function geodb_load_2018_us_county()::Dict{String,GeoRegion}
-    return load_census_gov(
+    return geodb_load_gov_census(
         db_name = "2018-us-county",
-        url = "https://www2.census.gov/geo/tiger/GENZ2018/shp/cb_2018_us_county_500k.zip",
-        expected_crc32 = 0x83eaec6d,
-        shp_filename = "cb_2018_us_county_500k.shp",
         extract_id = _id_2018_us_county,
+        shp_crc32 = 0x83eaec6d,
+        shp_filename = "cb_2018_us_county_500k.shp",
+        shp_url = "https://www2.census.gov/geo/tiger/GENZ2018/shp/cb_2018_us_county_500k.zip",
     )
 end
 
@@ -101,13 +111,13 @@ function _id_2018_us_zcta(table::Shapefile.Table, i::Int)::String
     return table.ZCTA5CE10[i]
 end
 
-function load_2018_us_zcta()::Dict{String,GeoPoint}
+function geodb_load_2018_us_zcta()::Dict{String,GeoRegion}
-    return load_census_gov(
+    return geodb_load_gov_census(
         db_name = "2018-us-zcta",
-        url = "https://www2.census.gov/geo/tiger/GENZ2018/shp/cb_2018_us_zcta510_500k.zip",
-        expected_crc32 = 0x6391f5fc,
-        shp_filename = "cb_2018_us_zcta510_500k.shp",
         extract_id = _id_2018_us_zcta,
+        shp_crc32 = 0x6391f5fc,
+        shp_filename = "cb_2018_us_zcta510_500k.shp",
+        shp_url = "https://www2.census.gov/geo/tiger/GENZ2018/shp/cb_2018_us_zcta510_500k.zip",
     )
 end
 
@@ -115,24 +125,24 @@ function _id_us_state(table::Shapefile.Table, i::Int)::String
     return table.STUSPS[i]
 end
 
-function load_us_state()::Dict{String,GeoPoint}
+function geodb_load_us_state()::Dict{String,GeoRegion}
-    return load_census_gov(
+    return geodb_load_gov_census(
         db_name = "us-state",
-        url = "https://www2.census.gov/geo/tiger/GENZ2018/shp/cb_2018_us_state_500k.zip",
-        expected_crc32 = 0x9469e5ca,
-        shp_filename = "cb_2018_us_state_500k.shp",
         extract_id = _id_us_state,
+        shp_crc32 = 0x9469e5ca,
+        shp_filename = "cb_2018_us_state_500k.shp",
+        shp_url = "https://www2.census.gov/geo/tiger/GENZ2018/shp/cb_2018_us_state_500k.zip",
     )
 end
 
-function load_latlon_database(db_name)
+function geodb_load(db_name::AbstractString)::Dict{String,GeoRegion}
-    db_name == "2018-us-county" && return load_2018_us_county()
+    db_name == "2018-us-county" && return geodb_load_2018_us_county()
-    db_name == "2018-us-zcta" && return load_2018_us_zcta()
+    db_name == "2018-us-zcta" && return geodb_load_2018_us_zcta()
-    db_name == "us-state" && return load_us_state()
+    db_name == "us-state" && return geodb_load_us_state()
     error("Unknown database: $db_name")
 end
 
-function geodb_query(name)
+function geodb_query(name)::GeoRegion
     db_name, id = split(name, ":")
-    return load_latlon_database(db_name)[id]
+    return geodb_load(db_name)[id]
 end

src/instance/parse.jl

@@ -54,9 +54,9 @@ function parse(json)::Instance
         if "initial amounts" in keys(product_dict)
             for (center_name, center_dict) in product_dict["initial amounts"]
                 if "location" in keys(center_dict)
-                    point = geodb_query(center_dict["location"])
+                    region = geodb_query(center_dict["location"])
-                    center_dict["latitude (deg)"] = point.lat
+                    center_dict["latitude (deg)"] = region.centroid.lat
-                    center_dict["longitude (deg)"] = point.lon
+                    center_dict["longitude (deg)"] = region.centroid.lon
                 end
                 center = CollectionCenter(
                     length(collection_centers) + 1,

test/instance/geodb_test.jl

@@ -5,19 +5,19 @@
 using RELOG
 
 @testset "geodb_query (2018-us-county)" begin
-    point = RELOG.geodb_query("2018-us-county:17043")
+    region = RELOG.geodb_query("2018-us-county:17043")
-    @test point.lat == 41.83956
+    @test region.centroid.lat == 41.83956
-    @test point.lon == -88.08857
+    @test region.centroid.lon == -88.08857
 end
 
 @testset "geodb_query (2018-us-zcta)" begin
-    point = RELOG.geodb_query("2018-us-zcta:60439")
+    region = RELOG.geodb_query("2018-us-zcta:60439")
-    @test point.lat == 41.68241
+    @test region.centroid.lat == 41.68241
-    @test point.lon == -87.98954
+    @test region.centroid.lon == -87.98954
 end
 
 @testset "geodb_query (us-state)" begin
-    point = RELOG.geodb_query("us-state:IL")
+    region = RELOG.geodb_query("us-state:IL")
-    @test point.lat == 39.73939
+    @test region.centroid.lat == 39.73939
-    @test point.lon == -89.50414
+    @test region.centroid.lon == -89.50414
 end