r10791 - /branches/multi_processor_merge/specific_fns/model_free/multi_processor_commands.py -- February 21, 2010 - 19:37

Author: bugman
Date: Sun Feb 21 19:37:31 2010
New Revision: 10791

URL: http://svn.gna.org/viewcvs/relax?rev=10791&view=rev
Log:
Deletion of the parallelised grid search classes.

The model-free multi-processor code is now far simpler.  However the 
parallelised grid search is not
functioning correctly yet.  The deleted classes are:
    MF_grid_memo
    MF_grid_result_command
    MF_split_grid_command


Modified:
    
branches/multi_processor_merge/specific_fns/model_free/multi_processor_commands.py

Modified: 
branches/multi_processor_merge/specific_fns/model_free/multi_processor_commands.py
URL: 
http://svn.gna.org/viewcvs/relax/branches/multi_processor_merge/specific_fns/model_free/multi_processor_commands.py?rev=10791&r1=10790&r2=10791&view=diff
==============================================================================
--- 
branches/multi_processor_merge/specific_fns/model_free/multi_processor_commands.py
 (original)
+++ 
branches/multi_processor_merge/specific_fns/model_free/multi_processor_commands.py
 Sun Feb 21 19:37:31 2010
@@ -54,101 +54,6 @@
     print("\n\n" + string)
     print(len(string) * '~')
 
-
-class MF_grid_memo(Memo):
-    def __init__(self, super_grid_memo):
-
-        # Execute the base class __init__() method.
-        super(MF_grid_memo, self).__init__()
-
-        self.super_grid_memo = super_grid_memo
-        self.super_grid_memo.add_sub_memo(self)
-
-
-    def add_result(self, results):
-        self.super_grid_memo.add_result(self, results)
-
-
-class MF_grid_result_command(Result_command):
-    def __init__(self, processor, result_string, memo_id, param_vector, 
func, iter, fc, gc, hc, warning, completed):
-
-        # Execute the base class __init__() method.
-        super(MF_grid_result_command, self).__init__(processor=processor, 
completed=completed)
-
-        self.result_string = result_string
-        self.memo_id = memo_id
-        self.param_vector = param_vector
-        self.func = func
-        self.iter = iter
-        self.fc = fc
-        self.gc = gc
-        self.hc = hc
-        self.warning = warning
-
-
-    def run(self, processor, memo):
-        # FIXME: Check against full result
-        # FIXME: names not consistent in memo
-        # FIXME: too much repacking
-        results = (self.param_vector, self.func, self.iter, self.fc, 
self.gc, self.hc, self.warning)
-        memo.add_result(results)
-
-        sgm = memo.super_grid_memo
-
-        print_prefix = sgm.print_prefix
-        verbosity = sgm.verbosity
-        full_output = sgm.full_output
-        A = sgm.A
-        b = sgm.b
-        grid_size = sgm.grid_size
-
-        if sgm.first_time:
-            print()
-            print("Unconstrained grid search size: " + repr(grid_size) + " 
(constraints may decrease this size).\n")
-            if verbosity:
-                if verbosity >= 2:
-                    print(print_prefix)
-                print(print_prefix)
-                print(print_prefix + "Grid search")
-                print(print_prefix + "~~~~~~~~~~~")
-
-            # Linear constraints.
-            if A != None and b != None:
-                if verbosity >= 3:
-                    print(print_prefix + "Linear constraint matrices.")
-                    print(print_prefix + "A: " + repr(A))
-                    print(print_prefix + "b: " + repr(b))
-
-        # we don't want to prepend the masters stdout tag
-        sys.stdout.write('\n'+self.result_string),
-
-        if sgm.completed:
-            if verbosity and results != None:
-                if full_output:
-                    print('')
-                    print('')
-                    print(print_prefix + "Parameter values: " + repr(sgm.xk))
-                    print(print_prefix + "Function value:   " + repr(sgm.fk))
-                    print(print_prefix + "Iterations:       " + repr(sgm.k))
-                    print(print_prefix + "Function calls:   " + 
repr(sgm.f_count))
-                    print(print_prefix + "Gradient calls:   " + 
repr(sgm.g_count))
-                    print(print_prefix + "Hessian calls:    " + 
repr(sgm.h_count))
-                    if sgm.warning:
-                        print(print_prefix + "Warning:          " + 
sgm.warning)
-                    else:
-                        print(print_prefix + "Warning:          None")
-                else:
-                    print(print_prefix + "Parameter values: " + 
repr(sgm.short_results))
-                print("")
-
-            # Initialise the iteration counter and function, gradient, and 
Hessian call counters.
-            sgm.model_free.iter_count = 0
-            sgm.model_free.f_count = 0
-            sgm.model_free.g_count = 0
-            sgm.model_free.h_count = 0
-
-            # Disassemble the results.
-            sgm.model_free._disassemble_result(param_vector=sgm.xk, 
func=sgm.fk, iter=sgm.k, fc=sgm.f_count, gc=sgm.g_count, hc=sgm.h_count, 
warning=sgm.warning, spin=sgm.spin, sim_index=sgm.sim_index, 
model_type=sgm.model_type, scaling=sgm.scaling, 
scaling_matrix=sgm.scaling_matrix)
 
 
 class MF_memo(Memo):
@@ -262,10 +167,6 @@
 
 
 
-class MF_split_grid_command(MF_minimise_command):
-    pass
-
-
 class MF_grid_command(MF_minimise_command):
     """Command class for the model-free grid search."""