Author: bugman
Date: Mon Nov 25 10:49:25 2013
New Revision: 21635
URL: http://svn.gna.org/viewcvs/relax?rev=21635&view=rev
Log:
Implemented model elimination for the relaxation dispersion analysis.
This currently uses the pA limits of 0.501 < pA < 0.999 to determine if a
model has failed.
To implement this, the dispersion API methods deselect(), eliminate(),
get_param_names() and
get_param_values() were written. These were copied from the model-free
analysis and modified as
needed.
Modified:
branches/relax_disp/specific_analyses/relax_disp/api.py
Modified: branches/relax_disp/specific_analyses/relax_disp/api.py
URL:
http://svn.gna.org/viewcvs/relax/branches/relax_disp/specific_analyses/relax_disp/api.py?rev=21635&r1=21634&r2=21635&view=diff
==============================================================================
--- branches/relax_disp/specific_analyses/relax_disp/api.py (original)
+++ branches/relax_disp/specific_analyses/relax_disp/api.py Mon Nov 25
10:49:25 2013
@@ -47,7 +47,7 @@
from specific_analyses.relax_disp.checks import check_c_modules,
check_disp_points, check_exp_type, check_exp_type_fixed_time,
check_model_type, check_pipe_type, check_spectra_id_setup
from specific_analyses.relax_disp.disp_data import average_intensity,
find_intensity_keys, get_curve_type, has_exponential_exp_type,
has_proton_mmq_cpmg, loop_cluster, loop_exp_frq_point,
loop_exp_frq_point_time, loop_frq, loop_time, pack_back_calc_r2eff,
return_cpmg_frqs, return_index_from_disp_point, return_index_from_exp_type,
return_index_from_frq, return_offset_data, return_param_key_from_data,
return_r1_data, return_r2eff_arrays, return_spin_lock_nu1,
spin_ids_to_containers
from specific_analyses.relax_disp.optimisation import Disp_memo,
Disp_minimise_command, back_calc_r2eff, grid_search_setup
-from specific_analyses.relax_disp.parameters import assemble_param_vector,
assemble_scaling_matrix, disassemble_param_vector, get_param_names,
linear_constraints, param_index_to_param_info, param_num
+from specific_analyses.relax_disp.parameters import assemble_param_vector,
assemble_scaling_matrix, disassemble_param_vector, get_param_names,
get_value, linear_constraints, loop_parameters, param_index_to_param_info,
param_num
from specific_analyses.relax_disp.variables import EXP_TYPE_CPMG_PROTON_MQ,
EXP_TYPE_CPMG_PROTON_SQ, MODEL_LIST_FULL, MODEL_LM63, MODEL_LM63_3SITE,
MODEL_CR72, MODEL_CR72_FULL, MODEL_DPL94, MODEL_IT99, MODEL_LIST_MMQ,
MODEL_M61, MODEL_M61B, MODEL_MMQ_2SITE, MODEL_MP05, MODEL_MQ_CR72,
MODEL_NOREX, MODEL_NS_CPMG_2SITE_3D, MODEL_NS_CPMG_2SITE_3D_FULL,
MODEL_NS_CPMG_2SITE_EXPANDED, MODEL_NS_CPMG_2SITE_STAR,
MODEL_NS_CPMG_2SITE_STAR_FULL, MODEL_NS_R1RHO_2SITE, MODEL_R2EFF,
MODEL_TAP03, MODEL_TP02, MODEL_TSMFK01
from target_functions.relax_disp import Dispersion
from user_functions.data import Uf_tables; uf_tables = Uf_tables()
@@ -787,6 +787,29 @@
default_value_doc.add_table(_table.label)
+ def deselect(self, model_info, sim_index=None):
+ """Deselect models or simulations.
+
+ @param model_info: The spin ID list from the model_loop() API
method.
+ @type model_info: int
+ @keyword sim_index: The optional Monte Carlo simulation index.
If None, then models will be deselected, otherwise the given simulation will.
+ @type sim_index: None or int
+ """
+
+ # Loop over all the spins and deselect them.
+ for spin_id in model_info:
+ # Get the spin.
+ spin = return_spin(spin_id)
+
+ # Spin deselection.
+ if sim_index == None:
+ spin.select = False
+
+ # Simulation deselection.
+ else:
+ spin.select_sim[sim_index] = False
+
+
def duplicate_data(self, pipe_from=None, pipe_to=None, model_info=None,
global_stats=False, verbose=True):
"""Duplicate the data specific to a single model.
@@ -871,6 +894,118 @@
# Duplicate the object.
new_obj = deepcopy(getattr(spin, name))
setattr(dp_to.mol[spin._mol_index].res[spin._res_index].spin[spin._spin_index],
name, new_obj)
+
+
+ def eliminate(self, name, value, model_info, args, sim=None):
+ """Relaxation dispersion model elimination, parameter by parameter.
+
+ @param name: The parameter name.
+ @type name: str
+ @param value: The parameter value.
+ @type value: float
+ @param model_info: The list of spin IDs from the model_loop() API
method.
+ @type model_info: int
+ @param args: The c1 and c2 elimination constant overrides.
+ @type args: None or tuple of float
+ @keyword sim: The Monte Carlo simulation index.
+ @type sim: int
+ @return: True if the model is to be eliminated, False
otherwise.
+ @rtype: bool
+ """
+
+ # Default limits.
+ c1 = 0.501
+ c2 = 0.999
+
+ # Depack the arguments.
+ if args != None:
+ c1, c2 = args
+
+ # The pA parameter.
+ if name == 'pA':
+ if value < c1:
+ if sim == None:
+ print("Data pipe '%s': The pA parameter of %.5f is less
than %.5f, eliminating the spin cluster %s." % (pipes.cdp_name(), value, c1,
model_info))
+ else:
+ print("Data pipe '%s': The pA parameter of %.5f is less
than %.5f, eliminating simulation %i of the spin cluster %s." %
(pipes.cdp_name(), value, c1, sim, model_info))
+ return True
+
+ if value > c2:
+ if sim == None:
+ print("Data pipe '%s': The pA parameter of %.5f is
greater than %.5f, eliminating the spin cluster %s." % (pipes.cdp_name(),
value, c1, model_info))
+ else:
+ print("Data pipe '%s': The pA parameter of %.5f is
greater than %.5f, eliminating simulation %i of the spin cluster %s." %
(pipes.cdp_name(), value, c1, sim, model_info))
+ return True
+
+ # Accept model.
+ return False
+
+
+ def get_param_names(self, model_info=None):
+ """Return a vector of parameter names.
+
+ @keyword model_info: The list spin ID strings from the
model_loop() API method.
+ @type model_info: list of str
+ @return: The vector of parameter names.
+ @rtype: list of str
+ """
+
+ # Get the spin containers.
+ spins = []
+ for spin_id in model_info:
+ # Get the spin.
+ spin = return_spin(spin_id)
+
+ # Skip deselected spins.
+ if not spin.select:
+ continue
+
+ # Add the spin.
+ spins.append(spin)
+
+ # No spins.
+ if not len(spins):
+ return None
+
+ # Call the get_param_names() function.
+ return get_param_names(spins)
+
+
+ def get_param_values(self, model_info=None, sim_index=None):
+ """Return a vector of parameter values.
+
+ @keyword model_info: The model index from model_info(). This is
zero for the global models or equal to the global spin index (which covers
the molecule, residue, and spin indices).
+ @type model_info: int
+ @keyword sim_index: The Monte Carlo simulation index.
+ @type sim_index: int
+ @return: The vector of parameter values.
+ @rtype: list of str
+ """
+
+ # Get the spin containers.
+ spins = []
+ for spin_id in model_info:
+ # Get the spin.
+ spin = return_spin(spin_id)
+
+ # Skip deselected spins.
+ if not spin.select:
+ continue
+
+ # Add the spin.
+ spins.append(spin)
+
+ # No spins.
+ if not len(spins):
+ return None
+
+ # Loop over the parameters of the cluster, fetching their values.
+ values = []
+ for param_name, param_index, spin_index, r20_key in
loop_parameters(spins=spins):
+ values.append(get_value(spins=spins, sim_index=sim_index,
param_name=param_name, spin_index=spin_index, r20_key=r20_key))
+
+ # Return all values.
+ return values
def grid_search(self, lower=None, upper=None, inc=None,
constraints=True, verbosity=1, sim_index=None):
r21635 - /branches/relax_disp/specific_analyses/relax_disp/api.py