replace graph helpers with np/builtins

graph/my_graph.py

@@ -5,7 +5,7 @@
 import warnings
 import json
 import my_graph_helpers as mgh
-from lazy_property import lazy_property
+from utils.lazy_property import lazy_property
 from matplotlib import pyplot as plt
 
 

graph/my_graph_helpers.py

@@ -168,32 +168,8 @@ def WeightedPick(d):
     """picks an item out of the dictionary d, with probability proportional to
     the value of that item.  e.g. in {a:1, b:0.6, c:0.4} selects and returns
     "a" 5/10 times, "b" 3/10 times and "c" 2/10 times. """
-
-    r = random.uniform(0, sum(d.values()))
-    s = 0.0
-    for k, w in d.items():
-        s += w
-        if r < s:
-            return k
-    return k
-
-
-def mat_reorder(matrix, order):
-    """sorts a square matrix so both rows and columns are
-    ordered by order. """
-
-    Drow = [matrix[i] for i in order]
-    Dcol = [[r[i] for i in order] for r in Drow]
-
-    return Dcol
-
-
-def myRoll(mylist):
-    """rolls a list, putting the last element into the first slot. """
-
-    mylist.insert(0, mylist[-1])
-    del mylist[-1]
-    return(mylist)
+    sum_p = sum(d.values())
+    return np.random.choice(d.keys(), p=[p/sum_p for p in d.values()])
 
 ######################
 # DUALS HElPER
@@ -467,8 +443,7 @@ def choose_path(myG, paths, alpha, strict_greedy=False):
     length more frequently  """
 
     if strict_greedy is False:
-        inv_weight = dict((k, 1.0/(paths[k]**alpha)) for k in paths)
-        target_path = WeightedPick(inv_weight)
+        target_path = WeightedPick({path: length**-alpha for (path, length) in paths.items()})
     if strict_greedy is True:
         target_path = min(paths, key=paths.get)
 

tests/perf_test_data

@@ -5,3 +5,17 @@
 1d11d71 NYC 1 0.142737150192 7.00588458333
 1d11d71 Prague 2 0.0788838863373 25.353720422
 1d11d71 CapeTown 3 63.7222030163 0.0470793515917
+e12af17 Phule_Nagar_v6 199 3.72162985802 53.4711961135
+e12af17 Epworth_Before 350 84.1490020752 4.15928877787
+e12af17 Epworth_demo 1 0.195746898651 5.10863777097
+e12af17 Las_Vegas 1 52.0890500546 0.0191978928192
+e12af17 NYC 1 0.135759830475 7.36594909188
+e12af17 Prague 2 0.0757038593292 26.4187323833
+e12af17 CapeTown 3 63.6627459526 0.0471233207916
+4a121ea Phule_Nagar_v6 199 3.75716304779 52.9654948345
+4a121ea Epworth_Before 350 85.5207681656 4.09257315512
+4a121ea Epworth_demo 1 0.192051172256 5.20694556691
+4a121ea Las_Vegas 1 51.1493070126 0.0195506070054
+4a121ea NYC 1 0.129373073578 7.72958369423
+4a121ea Prague 2 0.0711929798126 28.0926575242
+4a121ea CapeTown 3 56.6484811306 0.052958171872