Author: bugman
Date: Wed Oct 15 22:40:20 2008
New Revision: 7718
URL: http://svn.gna.org/viewcvs/relax?rev=7718&view=rev
Log:
Manually ported r3242 from the multi_processor branch.
The command used was:
svn merge -r3241:3242
svn+ssh://bugman@xxxxxxxxxxx/svn/relax/branches/multi_processor/specific_fns/model_free.py
specific_fns/model_free/mf_minimise.py
This was a bugger!
....
r3242 | varioustoxins | 2007-03-27 17:04:02 +0200 (Tue, 27 Mar 2007) | 3
lines
Changed paths:
M /branches/multi_processor/specific_fns/model_free.py
changes to add minimise command processing for multiprocessing
....
Modified:
branches/multi_processor_merge/specific_fns/model_free/mf_minimise.py
Modified:
branches/multi_processor_merge/specific_fns/model_free/mf_minimise.py
URL:
http://svn.gna.org/viewcvs/relax/branches/multi_processor_merge/specific_fns/model_free/mf_minimise.py?rev=7718&r1=7717&r2=7718&view=diff
==============================================================================
--- branches/multi_processor_merge/specific_fns/model_free/mf_minimise.py
(original)
+++ branches/multi_processor_merge/specific_fns/model_free/mf_minimise.py Wed
Oct 15 22:40:20 2008
@@ -25,6 +25,8 @@
from numpy import float64, array, dot, zeros
from numpy.linalg import inv
from re import match
+import cStringIO #FIXME: temporary for pickle tests
+import marshal #FIXME: temporary for pickle tests
# relax module imports.
from float import isNaN, isInf
@@ -33,6 +35,7 @@
from generic_fns.mol_res_spin import count_spins, exists_mol_res_spin_data,
return_spin_from_index, spin_loop
from maths_fns.mf import Mf
from minfx.generic import generic_minimise
+from multi.mpi4py_processor import MF_minimise_command
from physical_constants import h_bar, mu0, return_gyromagnetic_ratio
from relax_errors import RelaxError, RelaxInfError, RelaxLenError,
RelaxNaNError, RelaxNoModelError, RelaxNoPdbError, RelaxNoResError,
RelaxNoSequenceError, RelaxNoTensorError, RelaxNoValueError,
RelaxNoVectorsError, RelaxNucleusError, RelaxProtonTypeError,
RelaxSpinTypeError
@@ -938,12 +941,70 @@
relax_data, relax_error, equations, param_types, param_values,
r, csa, num_frq, frq, num_ri, remap_table, noe_r1_table, ri_labels, gx, gh,
num_params, xh_unit_vectors, diff_type, diff_params =
self.minimise_data_setup(model_type, min_algor, num_data_sets, min_options,
spin=spin, sim_index=sim_index)
- # Initialise the function to minimise.
- ######################################
-
- self.mf = Mf(init_params=param_vector, model_type=model_type,
diff_type=diff_type, diff_params=diff_params, scaling_matrix=scaling_matrix,
num_spins=num_spins, equations=equations, param_types=param_types,
param_values=param_values, relax_data=relax_data, errors=relax_error,
bond_length=r, csa=csa, num_frq=num_frq, frq=frq, num_ri=num_ri,
remap_table=remap_table, noe_r1_table=noe_r1_table, ri_labels=ri_labels,
gx=gx, gh=gh, h_bar=h_bar, mu0=mu0, num_params=num_params,
vectors=xh_unit_vectors)
-
-
+# # Initialise the function to minimise.
+# ######################################
+# print 'initialise Mf'
+# print 'init_params',self.param_vector
+# print 'param_se',self.param_set
+# print 'diff_type',diff_type
+# print 'diff_params',diff_params
+# print 'scaling_matrix',self.scaling_matrix
+# print 'num_res',num_res
+# print 'equations',equations
+# print 'param_types',param_types
+# print 'param_values',param_values
+# print 'relax_data',relax_data
+# print 'errors',relax_error
+# print 'bond_length',r
+# print 'csa=',csa
+# print 'num_frq',num_frq
+# print 'frq',frq
+# print 'num_ri',num_ri
+# print 'remap_table',remap_table
+# print 'noe_r1_table',noe_r1_table
+# print 'ri_labels',ri_labels
+# print 'gx',self.relax.data.gx
+# print 'gh',self.relax.data.gh
+# print 'g_ratio',self.relax.data.g_ratio
+# print 'h_bar',self.relax.data.h_bar
+# print 'mu0',self.relax.data.mu0
+# print 'num_params',num_params
+# print 'vectors',xh_unit_vectors
+#
+# data_list = [ 'initialise Mf',
+# 'init_params',self.param_vector,
+# 'param_set',self.param_set,
+# 'diff_type',diff_type,
+# 'diff_params',diff_params,
+# 'scaling_matrix',self.scaling_matrix,
+# 'num_res',num_res,
+# 'equations',equations,
+# 'param_types',param_types,
+# 'param_values',param_values,
+# 'relax_data',relax_data,
+# 'errors',relax_error,
+# 'bond_length',r,
+# 'csa=',csa,
+# 'num_frq',num_frq,
+# 'frq',frq,
+# 'num_ri',num_ri,
+# 'remap_table',remap_table,
+# 'noe_r1_table',noe_r1_table,
+# 'ri_labels',ri_labels,
+# 'gx',self.relax.data.gx,
+# 'gh',self.relax.data.gh,\
+# 'g_ratio',self.relax.data.g_ratio,
+# 'h_bar',self.relax.data.h_bar,
+# 'mu0',self.relax.data.mu0,
+# 'num_params',num_params,
+# 'vectors',xh_unit_vectors]
+# for elem in data_list:
+# marshal.loads(marshal.dumps(elem))
+# self.mf = Mf(init_params=param_vector, model_type=model_type,
diff_type=diff_type, diff_params=diff_params, scaling_matrix=scaling_matrix,
num_spins=num_spins, equations=equations, param_types=param_types,
param_values=param_values, relax_data=relax_data, errors=relax_error,
bond_length=r, csa=csa, num_frq=num_frq, frq=frq, num_ri=num_ri,
remap_table=remap_table, noe_r1_table=noe_r1_table, ri_labels=ri_labels,
gx=gx, gh=gh, h_bar=h_bar, mu0=mu0, num_params=num_params,
vectors=xh_unit_vectors)
+ command=MF_minimise_command()
+ command.set_mf(init_params=param_vector, model_type=model_type,
diff_type=diff_type, diff_params=diff_params, scaling_matrix=scaling_matrix,
num_spins=num_spins, equations=equations, param_types=param_types,
param_values=param_values, relax_data=relax_data, errors=relax_error,
bond_length=r, csa=csa, num_frq=num_frq, frq=frq, num_ri=num_ri,
remap_table=remap_table, noe_r1_table=noe_r1_table, ri_labels=ri_labels,
gx=gx, gh=gh, h_bar=h_bar, mu0=mu0, num_params=num_params,
vectors=xh_unit_vectors)
+ #test.assert_mf_equivalent(self.mf)
+ ##self.mf=test.mf
# Setup the minimisation algorithm when constraints are present.
################################################################
@@ -975,25 +1036,32 @@
# Back-calculation.
###################
+ #FIXME could be neater?
if min_algor == 'back_calc':
- return self.mf.calc_ri()
+ return command.build_mf().calc_ri()
# Minimisation.
###############
-
+ #FIXME remove old version
+# if constraints:
+# results = generic_minimise(func=self.mf.func,
dfunc=self.mf.dfunc, d2func=self.mf.d2func, args=(), x0=self.param_vector,
min_algor=min_algor, min_options=min_options, func_tol=func_tol,
grad_tol=grad_tol, maxiter=max_iterations, A=A, b=b, full_output=1,
print_flag=print_flag)
+# else:
+# results = generic_minimise(func=self.mf.func,
dfunc=self.mf.dfunc, d2func=self.mf.d2func, args=(), x0=self.param_vector,
min_algor=min_algor, min_options=min_options, func_tol=func_tol,
grad_tol=grad_tol, maxiter=max_iterations, full_output=1,
print_flag=print_flag)
+# if results == None:
+# return
+ #FIXME??? strange contraints
# Constrained optimisation.
if constraints:
- results = generic_minimise(func=self.mf.func,
dfunc=self.mf.dfunc, d2func=self.mf.d2func, args=(), x0=param_vector,
min_algor=min_algor, min_options=min_options, func_tol=func_tol,
grad_tol=grad_tol, maxiter=max_iterations, A=A, b=b, full_output=1,
print_flag=verbosity)
-
- # Unconstrained optimisation.
- else:
- results = generic_minimise(func=self.mf.func,
dfunc=self.mf.dfunc, d2func=self.mf.d2func, args=(), x0=param_vector,
min_algor=min_algor, min_options=min_options, func_tol=func_tol,
grad_tol=grad_tol, maxiter=max_iterations, full_output=1,
print_flag=verbosity)
+ command.set_minimise(args=(), x0=param_vector,
min_algor=min_algor, min_options=min_options, func_tol=func_tol,
grad_tol=grad_tol, maxiter=max_iterations, A=A, b=b, full_output=1,
print_flag=verbosity)
+ else:
+ command.set_minimise(args=(), x0=self.param_vector,
min_algor=min_algor, min_options=min_options, func_tol=func_tol,
grad_tol=grad_tol, maxiter=max_iterations, full_output=1,
print_flag=verbosity)
+ command.run(None)
# Disassemble the results.
if results == None:
continue
- param_vector, func, iter, fc, gc, hc, warning = results
+ param_vector, func, iter, fc, gc, hc, warning = command.results
iter_count = iter_count + iter
f_count = f_count + fc
g_count = g_count + gc
r7718 - /branches/multi_processor_merge/specific_fns/model_free/mf_minimise.py