Merge branch 'release/v0.3'

README.md

@@ -16,11 +16,6 @@ Since this is a new project there are still some obvious improvements which are
 ## Author
 Kristoffer Carlsson (@KristofferC)
 
-## Requirements
-
-Currently the "ArrayViews" package is required because `sub` in v0.3 of Julia
-is quite bad.
-
 ## Examples
 
 In the examples, notice that the module is called `KDtree` and the actual tree type is called `KDTree`. This is because modules and types can currently not have the same name in Julia.

benchmark/bench_build_tree.jl

@@ -1,27 +1,41 @@
 using KDtree
 
-dims = 1:6
-n_points = [10^i for i in 1:5]
-
-times = Array(Float64, length(dims), length(n_points))
-
-# Compile
-KDTree(randn(2,2))
-
-i = 0
-for dim in dims
-    j = 0
-    i+=1
-    for n_point in n_points
-        j+=1
-        data = randn(dim, n_point)
-        times[i,j]  = @elapsed KDTree(data)
+
+function run_bench_tree()
+    dims = 1:6
+    n_points = [10^i for i in 1:5]
+
+    times = fill(0.0, length(dims), length(n_points))
+
+    # Compile
+    KDTree(randn(2,2))
+
+    n_iters = 5
+
+    for (i, dim) in enumerate(dims)
+        for (j, n_point) in enumerate(n_points)
+            for z = 1:n_iters
+            	data = rand(dim, n_point)
+            	times[i,j] += @elapsed KDTree(data)
+        	end
+        	times[i,j] /= n_iters
+        end
     end
+
+    println(times)
 end
 
-println(times)
+run_bench_tree()
 
 #=
+2015-02-04: With type, no bounds, no views etc
+[[3.11e-6 3.2033e-5 0.000296701 0.003287041 0.08933542
+ 5.598e-6 3.2656e-5 0.000348329 0.004047143 0.049544467
+ 6.22e-6 3.6077e-5 0.000363257 0.02172358 0.058261703
+ 5.909e-6 3.8565e-5 0.000402444 0.004890908 0.080466724
+ 6.842e-6 4.0431e-5 0.000430746 0.005446367 0.093348975
+ 7.464e-6 4.3541e-5 0.000463091 0.005823931 0.103238088]
+
 2015-02-03: (with ArrayViews)
 [1.3373e-5 4.7584e-5 0.000489836 0.005438274 0.063624952
  7.775e-6 5.4737e-5 0.000640676 0.007319247 0.083582606

benchmark/bench_knn.jl

@@ -1,26 +1,47 @@
 using KDtree
 
-dims = 3
-n_points = [10^i for i in 3:6]
-ks = [1, 3, 10, 50, 100, 500]
 
-times = Array(Float64, length(ks), length(n_points))
+function run_bench_knn()
+    dims = 3
+    n_points = [10^i for i in 3:5]
+    ks = [1, 3, 10, 50, 100, 500]
 
-# Compile it
-tree = KDTree(randn(2,2))
-k_nearest_neighbour(tree, zeros(2), 1)
+    n_iters = 10
 
-for (i, k) in enumerate(ks)
-    for (j , n_point) in enumerate(n_points)
-        data = randn(dims, n_point)
-        tree = KDTree(data)
-        times[i,j]  = @elapsed k_nearest_neighbour(tree, zeros(dims), k)
+    times = fill(0.0, length(ks), length(n_points))
+
+    # Compile it
+    tree = KDTree(randn(2,2))
+    k_nearest_neighbour(tree, zeros(2), 1)
+
+    for (i, k) in enumerate(ks)
+        for (j , n_point) in enumerate(n_points)
+            data = rand(dims, n_point)
+            tree = KDTree(data)
+            for z in 1:n_iters
+            	p = rand(dims)
+                times[i,j]  += @elapsed k_nearest_neighbour(tree, p, k)
+            end
+            times[i,j] /= n_iters
+        end
     end
+
+    println(times)
+    return
 end
 
-println(times)
+run_bench_knn()
 
 #=
+2015-02-04 No bounds, type fixes
+[4.139509999999997e-6 6.266810000000003e-6 1.1824519999999996e-5 1.2813520000000002e-5
+ 6.926150000000001e-6 8.745540000000002e-6 1.5848950000000007e-5 1.9322909999999995e-5
+ 1.2340789999999995e-5 1.7301359999999997e-5 2.9695020000000008e-5 3.380342999999999e-5
+ 3.891018e-5 5.718188999999997e-5 8.192254999999999e-5 0.00010137918999999997
+ 6.649346000000001e-5 9.620713999999998e-5 0.00014474611999999997 0.00016459152000000002
+ 0.0004432140299999998 0.0006229702799999999 0.0008457700199999998 0.0009439302800000001]
+
+
 2015-02-03: ArrayViews:
 [1.3996e-5 2.1771e-5 5.1316e-5 4.4474e-5
  2.0837e-5 2.5502e-5 5.9402e-5 6.8732e-5

benchmark/bench_query_ball.jl

@@ -1,27 +1,40 @@
 using KDtree
 
-dims = 3
-n_points = [10^i for i in 3:5]
-rs = [0.1 0.2 0.3 0.4]
 
-times = Array(Float64, length(rs), length(n_points))
-
-# Compile it
-tree = KDTree(randn(2,2))
-query_ball_point(tree, zeros(2), 0.1)
-
-
-for (i, r) in enumerate(rs)
-    for (j , n_point) in enumerate(n_points)
-        data = randn(dims, n_point)
-        tree = KDTree(data)
-        times[i,j]  = @elapsed query_ball_point(tree, zeros(dims), r)
+function run_bench_query_ball()
+    dims = 3
+    n_points = [10^i for i in 3:5]
+    rs = [0.1 0.2 0.3 0.4]
+
+    times = fill(0.0, length(rs), length(n_points))
+
+    # Compile it
+    n_iters = 1000
+
+    tree = KDTree(randn(2,2))
+    query_ball_point(tree, zeros(2), 0.1)
+
+    for (i, r) in enumerate(rs)
+        for (j , n_point) in enumerate(n_points)
+            data = rand(dims, n_point)
+            tree = KDTree(data)
+            for z = 1:n_iters
+            	p = rand(dims)
+            	times[i,j] += @elapsed query_ball_point(tree, p, r)
+            end
+            times[i,j] /= n_iters
+        end
     end
+
+    println(times)
 end
 
-println(times)
+run_bench_query_ball()
+
 
 #=
+
+
 2015-02-03: ArrayViews + no sqrt
 [1.1196e-5 1.9593e-5 7.5885e-5
  1.5239e-5 4.1986e-5 0.000167011

src/KDtree.jl

@@ -1,5 +1,7 @@
 module KDtree
 
+    import Base.show
+
     using ArrayViews
 
     export KDTree