Simulator: implement own random generator instead of scipy

Dependency on scipy results in extra requirements to the system
(include the need of Fortran compiler). With this patch the
weighted random choice algorithm is implemented directly.

Change-Id: I60cbc6c452945000add5a65263d1ab4e42dd91f9

tools/simulator.py

@@ -14,6 +14,7 @@ import eventlet
 eventlet.monkey_patch()
 
 import argparse
+import bisect
 import collections
 import datetime
 import itertools
@@ -27,8 +28,6 @@ import threading
 import time
 import yaml
 
-from scipy.stats import rv_discrete
-
 from oslo_config import cfg
 import oslo_messaging as messaging
 from oslo_messaging import notify  # noqa
@@ -36,7 +35,7 @@ from oslo_messaging import rpc  # noqa
 from oslo_utils import timeutils
 
 LOG = logging.getLogger()
-RANDOM_VARIABLE = None
+RANDOM_GENERATOR = None
 CURRENT_PID = None
 RPC_CLIENTS = []
 MESSAGES = []
@@ -51,6 +50,8 @@ Usage example:
  --url rabbit://stackrabbit:secretrabbit@localhost/ rpc-client\
  --exit-wait 15000 -p 64 -m 64"""
 
+DISTRIBUTION_BUCKET_SIZE = 500
+
 
 def init_random_generator():
     data = []
@@ -61,18 +62,26 @@ def init_random_generator():
 
     ranges = collections.defaultdict(int)
     for msg_length in data:
-        range_start = (msg_length / 500) * 500 + 1
+        range_start = ((msg_length / DISTRIBUTION_BUCKET_SIZE) *
+                       DISTRIBUTION_BUCKET_SIZE + 1)
         ranges[range_start] += 1
 
     ranges_start = sorted(ranges.keys())
     total_count = len(data)
-    ranges_dist = []
-    for r in ranges_start:
-        r_dist = float(ranges[r]) / total_count
-        ranges_dist.append(r_dist)
 
-    random_var = rv_discrete(values=(ranges_start, ranges_dist))
-    return random_var
+    accumulated_distribution = []
+    running_total = 0
+    for range_start in ranges_start:
+        norm = float(ranges[range_start]) / total_count
+        running_total += norm
+        accumulated_distribution.append(running_total)
+
+    def weighted_random_choice():
+        r = random.random() * running_total
+        start = ranges_start[bisect.bisect_right(accumulated_distribution, r)]
+        return random.randrange(start, start + DISTRIBUTION_BUCKET_SIZE)
+
+    return weighted_random_choice
 
 
 class LoggingNoParsingFilter(logging.Filter):
@@ -214,10 +223,9 @@ def init_msg(messages_count):
     if messages_count > 1000:
         messages_count = 1000
     LOG.info("Preparing %d messages", messages_count)
-    ranges = RANDOM_VARIABLE.rvs(size=messages_count)
-    i = 0
-    for range_start in ranges:
-        length = random.randint(range_start, range_start + 497)
+
+    for i in range(messages_count):
+        length = RANDOM_GENERATOR()
         msg = ''.join(random.choice(string.lowercase) for x in range(length)) \
               + ' ' + str(i)
         MESSAGES.append(msg)
@@ -472,6 +480,6 @@ def main():
 
 
 if __name__ == '__main__':
-    RANDOM_VARIABLE = init_random_generator()
+    RANDOM_GENERATOR = init_random_generator()
     CURRENT_PID = os.getpid()
     main()