# HG changeset patch # User paulb # Date 1189971362 0 # Node ID 2318878da974e4b4b627c15a3fba56c6c8a096d4 # Parent 9abc49db3422ae09d2fad7ae0a68f83577e57e93 [project @ 2007-09-16 19:36:02 by paulb] Added MakeReusable and reuse documentation. Improved the summary table. diff -r 9abc49db3422 -r 2318878da974 docs/tutorial.xhtml --- a/docs/tutorial.xhtml Sun Sep 16 19:35:34 2007 +0000 +++ b/docs/tutorial.xhtml Sun Sep 16 19:36:02 2007 +0000 @@ -159,7 +159,7 @@ words - we can make this code use several processes instead of just one. Here is the modified code:

+     t = time.time()
 
     # Initialise the results using a map with a limit on the number of
@@ -525,6 +525,58 @@
 the results in the same order as the initiation of the computations which
 produced them.
 
+Reusing Processes in Parallel Programs
+
+One notable aspect of the above programs when parallelised is that each
+invocation of a computation in parallel creates a new process in which the
+computation is to be performed, regardless of whether existing processes had
+just finished producing results and could theoretically have been asked to
+perform new computations. In other words, processes were created and destroyed
+instead of being reused.
+
+However, we can request that processes be reused for computations by
+enabling the reuse feature of exchange-like objects and employing
+suitable reusable callables. Consider this modified version of the simple_managed_map program:
+
++    t = time.time()
+
+    # Initialise the results using a map with a limit on the number of
+    # channels/processes.
+
+    results = pprocess.Map(limit=limit, reuse=1)
+
+    # Wrap the calculate function and manage it.
+
+    calc = results.manage(pprocess.MakeReusable(calculate))
+
+    # Perform the work.
+
+    print "Calculating..."
+    for i in range(0, N):
+        for j in range(0, N):
+            calc(i, j)
+
+    # Show the results.
+
+    for i in range(0, N):
+        for result in results[i*N:i*N+N]:
+            print result,
+        print
+
+    print "Time taken:", time.time() - t
+
+
+(This code in context with import statements and functions is
+found in the examples/simple_manage_map_reusable.py file.)
+
+By indicating that processes and channels shall be reused, and by wrapping
+the calculate function with the necessary support, the
+computations may be performed in parallel using a pool of processes instead of
+creating a new process for each computation and then discarding it, only to
+create a new process for the next computation.
+
 Summary
 
 The following table indicates the features used in converting one
@@ -547,16 +599,16 @@
     
       simple1
       simple_managed_map
-      Map, manage
+      MakeParallel, Map, manage
     
     
       simple2
       simple_managed_queue
-      Queue, manage
+      MakeParallel, Queue, manage
     
     
       simple_managed
-      Exchange (subclass), manage, finish
+      MakeParallel, Exchange (subclass), manage, finish
     
     
       simple_start