Merge pull request #26 from IntelPython/example-work

Improving parallel MC use-case example

Author: samaid

examples/README.md

@@ -17,10 +17,40 @@ This data is post-processed as necessary for the application.
 Code is tested to estimate the probability that 3 segments, obtained by splitting a unit stick 
 in two randomly chosen places, can be sides of a triangle. This probability is known in closed form to be $\frac{1}{4}$.
 
+Run python script "stick_triangle.py" to estimate this probability using parallel Monte-Carlo algorithm:
+
+```
+> python stick_triangle.py
+Parallel Monte-Carlo estimation of stick triangle probability
+Parameters: n_workers=12, batch_size=262144, n_batches=10000, seed=77777
+
+Monte-Carlo estimate of probability: 0.250000
+Population estimate of the estimator's standard deviation: 0.000834
+Expected standard deviation of the estimator: 0.000846
+Execution time: 64.043 seconds
+
+```
+
 ## Stick tetrahedron problem
 
 Code is used to estimate the probability that 6 segments, obtained by splitting a unit stick in 
 5 random chosen places, can be sides of a tetrahedron. 
 
 The probability is not known in closed form. See
-[math.stackexchange.com/questions/351913](https://math.stackexchange.com/questions/351913/probability-that-a-stick-randomly-broken-in-five-places-can-form-a-tetrahedron) for more details.
+[math.stackexchange.com/questions/351913](https://math.stackexchange.com/questions/351913/probability-that-a-stick-randomly-broken-in-five-places-can-form-a-tetrahedron) for more details.
+
+```
+> python stick_tetrahedron.py -s 1274 -p 4 -n 8096
+Parallel Monte-Carlo estimation of stick tetrahedron probability
+Input parameters: -s 1274 -b 65536 -n 8096 -p 4 -d 0
+
+Monte-Carlo estimate of probability: 0.01257113
+Population estimate of the estimator's standard deviation: 0.00000488
+Expected standard deviation of the estimator: 0.00000484
+Total MC size: 530579456
+
+Bayesian posterior beta distribution parameters: (6669984, 523909472)
+
+Execution time: 30.697 seconds
+
+```

examples/parallel_mc.py

@@ -1,21 +1,23 @@
 import multiprocessing as mp
+from functools import partial
 
 __all__ = ['parallel_mc_run']
 
 def worker_compute(w_id):
     "Worker function executed on the spawned slave process"
-    # global _local_rs
-    return _worker_mc_compute(_local_rs)
+    global _local_rs, _worker_mc_compute_func
+    return _worker_mc_compute_func(_local_rs)
 
 
-def assign_worker_rs(w_rs):
+def init_worker(w_rs, mc_compute_func=None, barrier=None):
     """Assign process local random state variable `rs` the given value"""
     assert not '_local_rs' in globals(), "Here comes trouble. Process is not expected to have global variable `_local_rs`"
 
-    global _local_rs
+    global _local_rs, _worker_mc_compute_func
     _local_rs = w_rs
+    _worker_mc_compute_func = mc_compute_func
     # wait to ensure that the assignment takes place for each worker
-    b.wait()
+    barrier.wait()
 
 def parallel_mc_run(random_states, n_workers, n_batches, mc_func):
     """
@@ -25,15 +27,15 @@ def parallel_mc_run(random_states, n_workers, n_batches, mc_func):
     and has access to worker-local global variable `rs`, containing worker's random states.
     """
     # use of Barrier ensures that every worker gets one
-    global b, _worker_mc_compute
-    b = mp.Barrier(n_workers)
+
+    with mp.Manager() as manager:
+        b = manager.Barrier(n_workers)
 
-    _worker_mc_compute = mc_func
-    with mp.Pool(processes=n_workers) as pool:
-        # 1. map over every worker once to distribute RandomState instances
-        pool.map(assign_worker_rs, random_states, chunksize=1)
-        # 2. Perform computations on workers
-        r = pool.map(worker_compute, range(n_batches), chunksize=1)
+        with mp.Pool(processes=n_workers) as pool:
+            # 1. map over every worker once to distribute RandomState instances
+            pool.map(partial(init_worker, mc_compute_func=mc_func, barrier=b), random_states, chunksize=1)
+            # 2. Perform computations on workers
+            r = pool.map(worker_compute, range(n_batches), chunksize=1)
 
     return r
 

examples/stick_tetrahedron.py

@@ -4,7 +4,7 @@
 from sticky_math import mc_six_piece_stick_tetrahedron_prob
 from arg_parsing import parse_arguments
 
-def mc_runner(rs):
+def mc_runner(rs, batch_size=None):
     return mc_six_piece_stick_tetrahedron_prob(rs, batch_size)
 
 def aggregate_mc_counts(counts, n_batches, batch_size):
@@ -40,6 +40,7 @@ def print_result(p_est, p_std, mc_size):
     from itertools import repeat
     from timeit import default_timer as timer
     import sys
+    from functools import partial
 
     args = parse_arguments()
 
@@ -51,23 +52,24 @@ def print_result(p_est, p_std, mc_size):
     batch_size = args.batch_size
     batches = args.batch_count
     id0 = args.id_offset
+    print("Parallel Monte-Carlo estimation of stick tetrahedron probability")
+    print("Input parameters: -s {seed} -b {batchSize} -n {numBatches} -p {processes} -d {idOffset}".format(
+        seed=args.seed, batchSize=args.batch_size, numBatches=args.batch_count, processes=n_workers, idOffset=args.id_offset))
+    print("")
 
     t0 = timer()
 
     rss = build_MT2203_random_states(seed, id0, n_workers)
-    r = parallel_mc_run(rss, n_workers, batches, mc_runner)
-    # r = sequential_mc_run(rss, n_workers, batches, mc_runner)
+
+    r = parallel_mc_run(rss, n_workers, batches, partial(mc_runner, batch_size=batch_size))
+    # r = sequential_mc_run(rss, n_workers, batches, partial(mc_runner, batch_size=batch_size))
 
     # retrieve values of estimates into numpy array
     counts = np.fromiter(r, dtype=np.double)
     p_est, p_std, event_count, nonevent_count = aggregate_mc_counts(counts, batches, batch_size)
 
     t1 = timer()
 
-
-    print("Input parameters: -s {seed} -b {batchSize} -n {numBatches} -p {processes} -d {idOffset}".format(
-        seed=args.seed, batchSize=args.batch_size, numBatches=args.batch_count, processes=n_workers, idOffset=args.id_offset))
-    print("")
     print_result(p_est, p_std, batches * batch_size)
     print("")
     print("Bayesian posterior beta distribution parameters: ({0}, {1})".format(event_count, nonevent_count))

examples/fancy.py renamed to examples/stick_triangle.py

@@ -39,41 +39,46 @@ def mc_dist(rs, n):
     return mc_prob
 
 
-def assign_worker_rs(w_rs):
+def init_worker(w_rs, barrier=None):
     """Assign process local random state variable `rs` the given value"""
     assert not 'rs' in globals(), "Here comes trouble. Process is not expected to have global variable `rs`"
 
     global rs
     rs = w_rs
     # wait to ensure that the assignment takes place for each worker
-    b.wait()
+    barrier.wait()
 
 
-def worker_compute(w_id):
+def worker_compute(w_id, batch_size=None):
     return mc_dist(rs, batch_size)
 
 
 if __name__ == '__main__':
     import multiprocessing as mp
     from itertools import repeat
     from timeit import default_timer as timer
+    from functools import partial
 
     seed = 77777
     n_workers = 12
     batch_size = 1024 * 256
     batches = 10000
+    print("Parallel Monte-Carlo estimation of stick triangle probability")
+    print(f"Parameters: n_workers={n_workers}, batch_size={batch_size}, n_batches={batches}, seed={seed}")
+    print("")
 
     t0 = timer()
     # Create instances of RandomState for each worker process from MT2203 family of generators
     rss = [ rnd.RandomState(seed, brng=('MT2203', idx)) for idx in range(n_workers) ]
-    # use of Barrier ensures that every worker gets one
-    b = mp.Barrier(n_workers)
-
-    with mp.Pool(processes=n_workers) as pool:
-        # map over every worker once to distribute RandomState instances
-        pool.map(assign_worker_rs, rss, chunksize=1)
-        # Perform computations on workers
-        r = pool.map(worker_compute, range(batches), chunksize=1)
+    with mp.Manager() as manager:
+        # use of Barrier ensures that every worker gets one
+        b = manager.Barrier(n_workers)
+
+        with mp.Pool(processes=n_workers) as pool:
+            # map over every worker once to distribute RandomState instances
+            pool.map(partial(init_worker, barrier=b), rss, chunksize=1)
+            # Perform computations on workers
+            r = pool.map(partial(worker_compute, batch_size=batch_size), range(batches), chunksize=1)
 
     # retrieve values of estimates into numpy array
     ps = np.fromiter(r, dtype=np.double)