From: Tomas Musil <tomik.musil@gmail.com>
Date: Tue, 10 Dec 2013 15:39:05 +0000 (+0100)
Subject: restarting the search
X-Git-Url: http://git.tomasm.cz/imago.git/commitdiff_plain/676b17ccc2d334fd2574bd26aef14c8381f6aaaa

restarting the search
---

diff --git a/imago_pack/cs.py b/imago_pack/cs.py
index 7c91119..6273651 100644
--- a/imago_pack/cs.py
+++ b/imago_pack/cs.py
@@ -14,10 +14,10 @@ class Space(object):
         self.best_value, self.best = max(self.nests)
 
 def new_nest(space):
-        position = [2 * space.bound * random.random() 
-                    - space.bound for _ in xrange(space.dimension)]
-        value = space.d_function(*position)
-        return (value, position)
+    position = [2 * space.bound * random.random() 
+                - space.bound for _ in xrange(space.dimension)]
+    value = space.d_function(*position)
+    return (value, position)
 
 def get_cuckoos(space):
     beta = 1.5
@@ -64,8 +64,13 @@ def next_turn(space):
     space.nests = [max(n, m) for (n, m) in zip(space.nests, nests)]
     space.best_value, space.best = max(space.nests)
 
-def optimize(dimension, boundary, function_d, n_nest, n_turns):
-    space = Space(dimension, boundary, function_d, n_nest)
-    for _ in xrange(n_turns):
-        next_turn(space)
-    return space.best
+def optimize(dimension, boundary, function_d, n_nest, n_turns, reset=1):
+    best_list = []
+    for i in xrange(reset):
+        space = Space(dimension, boundary, function_d, n_nest)
+        for _ in xrange(n_turns / reset):
+            next_turn(space)
+        best_list.append((space.best_value, space.best))
+        print space.best_value
+
+    return max(best_list)[1]
diff --git a/imago_pack/gridf.py b/imago_pack/gridf.py
index 3bb7c54..6a80c44 100644
--- a/imago_pack/gridf.py
+++ b/imago_pack/gridf.py
@@ -62,7 +62,7 @@ def find(lines, size, l1, l2, bounds, hough, show_all, do_something):
     f_dist = partial(job_4, im_l=im_l_s, v1=v1, v2=v2, h1=h1, h2=h2,
                      dv=delta_v, dh=delta_h, size=size)
 
-    x_v, y_v, x_h, y_h = Optimizer.optimize(4, 30, f_dist, 128, 256)
+    x_v, y_v, x_h, y_h = Optimizer.optimize(4, 30, f_dist, 64, 256, 4)
 
     v1 = (v1[0] + x_v * delta_v, v1[1] + x_v)
     v2 = (v2[0] + y_v * delta_v, v2[1] + y_v)