upd

config/Attractiveness.csv

@@ -1,4 +1,4 @@
-class	key	points	range	values
+class	key	influence	range	values
 education	amenity	3	2000	kindergarten
 education	amenity	5	3000	school,music_school,language_school
 education	amenity	20	10000	university,college

src/attractiveness.jl

@@ -4,8 +4,8 @@
 
 struct AttractivenessData
     class::Symbol
-    points::Int
-    range::Int
+    influence::Float64
+    range::Float64
     enu::ENU
     lla::LLA
 end
@@ -26,7 +26,7 @@ The CSV file or DataFrame should have the following columns:
     - class - data class in attractiveness index, each class name creates attractiveness dimension
     - key - key in the XML file <tag>
     - values - values in the <tag> (a star `"*"` catches all values)
-    - points - number of influance points 
+    - influence - strength of influence  
     - range - maximum influence range in meters 
 
 When a `DataFrame` is provided the additional parameter `refLLA` can be provided for the reference 
@@ -40,7 +40,7 @@ function AttractivenessSpatIndex(df::AbstractDataFrame, refLLA::LLA = LLA(mean(d
         enu = ENU(lla, refLLA)
         range_ = df.range[id]
         rect = SpatialIndexing.Rect((enu.east-range_,enu.north-range_), (enu.east+range_,enu.north+range_))
-        a = AttractivenessData(Symbol(df.class[id]), df.points[id], df.range[id], enu, lla)
+        a = AttractivenessData(Symbol(df.class[id]), df.influence[id], df.range[id], enu, lla)
         push!(data, SpatialElem(rect, id, a))
     end
     tree = RTree{Float64, 2}(Int, AttractivenessData, variant=SpatialIndexing.RTreeStar)
@@ -95,9 +95,9 @@ function attractiveness(sindex::AttractivenessSpatIndex, enu::ENU, aggregator::F
         a = item.val  # typeof(a) === AttractivenessData
         poidistance = OpenStreetMapX.distance(enu, a.enu)
         poidistance > a.range && continue
-        res[a.class] = aggregator(res[a.class],   a.points * (a.range - poidistance) / a.range)
+        res[a.class] = aggregator(res[a.class],   a.influence * (a.range - poidistance) / a.range)
         if explain
-            append!(explanation, DataFrame(;a.class,a.points,poidistance,a.lla.lat,a.lla.lon))
+            append!(explanation, DataFrame(;a.class,a.influence,poidistance,a.lla.lat,a.lla.lon))
         end
     end
     if explain

src/poi.jl

@@ -1,15 +1,15 @@
 
 struct Attract
     class::String
-    points::Int
+    influence::Int
     range::Int
 end
 """
     Represents the configuration of the data scraping process from OSM XML.
 Only those pieces of data will be scraped that are defined here.
 
 The configuration is defined in a DataFrame with the following columns:
-`class`, `key`, `values`, `points`, `range`.
+`class`, `key`, `values`, `influence`, `range`.
 Instead of the DataFrame a paths to a CSV file can be provided.
 
 * Constructors *
@@ -28,17 +28,17 @@ end
 const __builtin_attract_path = joinpath(@__DIR__, "..", "config", "Attractiveness.csv")
 
 function AttractivenessConfig(df::DataFrame)
-    colnames = ["class", "key", "points", "range", "values"]
+    colnames = ["class", "key", "influence", "range", "values"]
     @assert all(colnames .∈ Ref(names(df)))
-    dkeys = Set(df.key)
+    dkeys = Set(String.(df.key))
     attract = Dict{Union{String, Tuple{String,String}}, Attract}()
     for row in eachrow(df)
-        a = Attract(row.class, row.points, row.range)
-        for value in string.(split(row.values,','))
+        a = Attract(String(row.class), Float64(row.influence), Float64(row.range))  
+        for value in string.(split(String(row.values),','))
             if value == "*"
-                attract[row.key] = a
+                attract[String(row.key)] = a
             else
-                attract[row.key, value] = a
+                attract[String(row.key), value] = a
             end
         end
     end
@@ -47,7 +47,7 @@ end
 
 function AttractivenessConfig(filename::AbstractString = __builtin_attract_path)
     AttractivenessConfig(CSV.read(filename, DataFrame,types=Dict(
-        :class => String, :key => String, :points => Int, :range => Int, :values =>String) ))
+        :class => String, :key => String, :influence => Float64, :range => Float64, :values =>String) ))
 end
 
 const __builtin_attract = AttractivenessConfig()
@@ -156,12 +156,12 @@ function find_poi(filename::AbstractString; attract_config::AttractivenessConfig
                 a = get(attract, key, get(attract, (key, value), nothing))
                 if !isnothing(a)
                     # we are interested only in attractive POIs
-                    push!(df, (;elemtype,elemid,nodeid=curnode.id, lat=curnode.lat, lon=curnode.lon, key, value,a.class, a.points, a.range ) )
+                    push!(df, (;elemtype,elemid,nodeid=curnode.id, lat=curnode.lat, lon=curnode.lon, key, value,a.class, a.influence, a.range ) )
                 end
             end
         end
     end
-    df2 = DataFrame(g[findmax(g.points)[2], :]  for g in  groupby(df, [:nodeid, :class]))
+    df2 = DataFrame(g[findmax(g.influence)[2], :]  for g in  groupby(df, [:nodeid, :class]))
     df2
 end
 
@@ -176,11 +176,11 @@ function find_poi(osm::OSMData; attract_config::AttractivenessConfig=__builtin_a
         # otherwise try to get attractiveness for the tuple
         a = get(attract, key, get(attract, (key, value), nothing))
         if a !== nothing
-            # we are interested only in attractive POIs            #push!(df, (;elemtype, elemid, nodeid=curnode.id, lat=curnode.lat, lon=curnode.lon, key, value, a.class, a.points, a.range))
+            # we are interested only in attractive POIs
             lla = osm.nodes[node]
-            push!(df, (;nodeid=node, lat=lla.lat, lon=lla.lon, key, value, a.class, a.points, a.range))
+            push!(df, (;nodeid=node, lat=lla.lat, lon=lla.lon, key, value, a.class, a.influence, a.range))
         end
     end
-    df2 = DataFrame(g[findmax(g.points)[2], :] for g in groupby(df, [:nodeid, :class]))
+    df2 = DataFrame(g[findmax(g.influence)[2], :] for g in groupby(df, [:nodeid, :class]))
     return df2
 end

test/data/Attractiveness.csv

@@ -1,4 +1,4 @@
-class	key	points	range	values
+class	key	influence	range	values
 education	amenity	3	2000	kindergarten
 education	amenity	5	3000	school,music_school,language_school
 education	amenity	20	10000	university,college