Author: bugman
Date: Sun Apr 13 21:29:19 2008
New Revision: 5657
URL: http://svn.gna.org/viewcvs/relax?rev=5657&view=rev
Log:
Converted the whole of the select() function to the new relax design.
Modified:
1.3/generic_fns/model_selection.py
Modified: 1.3/generic_fns/model_selection.py
URL:
http://svn.gna.org/viewcvs/relax/1.3/generic_fns/model_selection.py?rev=5657&r1=5656&r2=5657&view=diff
==============================================================================
--- 1.3/generic_fns/model_selection.py (original)
+++ 1.3/generic_fns/model_selection.py Sun Apr 13 21:29:19 2008
@@ -27,6 +27,8 @@
# relax module imports.
from data import Data as relax_data_store
from relax_errors import RelaxDiffSeqError, RelaxError, RelaxNoPipeError,
RelaxNoSequenceError
+import pipes
+from specific_fns.setup import get_specific_fn
def aic(chi2, k, n):
@@ -94,135 +96,116 @@
return chi2 + k * log(n)
-def select(method=None, modsel_run=None, runs=None):
+def select(method=None, pipes=None):
"""Model selection function."""
- # Test if the model selection run exists.
- if not modsel_run in relax_data_store.run_names:
- raise RelaxNoPipeError, modsel_run
-
- # The runs argument.
- if runs == None:
- # Use the runs from 'relax_data_store.run_names'.
- self.runs = deepcopy(relax_data_store.run_names)
- # Remove the model selection run name if it is in the list.
- if modsel_run in self.runs:
- self.runs.remove(modsel_run)
- else:
- self.runs = runs
+ # Use all pipes (but the current).
+ if pipes == None:
+ # Get all data pipe names from the relax data store.
+ pipes = relax_data_store.keys()
+
+ # Remove the current data pipe.
+ pipes.remove(relax_data_store.current_pipe)
+
+ # Store the current data pipe as the model selection run
+ modsel_pipe = relax_data_store.current_pipe
# Select the model selection technique.
if method == 'AIC':
print "AIC model selection."
- self.formula = self.aic
+ formula = aic
elif method == 'AICc':
print "AICc model selection."
- self.formula = self.aicc
+ formula = aicc
elif method == 'BIC':
print "BIC model selection."
- self.formula = self.bic
+ formula = bic
elif method == 'CV':
print "CV model selection."
raise RelaxError, "The model selection technique " + `method` + " is
not currently supported."
else:
raise RelaxError, "The model selection technique " + `method` + " is
not currently supported."
- # No runs.
- if len(self.runs) == 0:
- raise RelaxError, "No runs are availible for use in model selection."
+ # No pipes.
+ if len(pipes) == 0:
+ raise RelaxError, "No data pipes are available for use in model
selection."
# Initialise.
- self.first_run = None
- self.function_type = {}
- self.duplicate_data = {}
- self.count_num_instances = {}
- self.model_statistics = {}
- self.skip_function = {}
+ first_run = None
+ function_type = {}
+ count_num_instances = {}
+ model_statistics = {}
+ skip_function = {}
# Cross validation setup.
- if type(self.runs[0]) == list:
- # No runs.
- if len(run) == 0:
- raise RelaxError, "No runs are availible for use in model
selection in the array " + `run` + "."
-
- # Loop over the runs.
- for i in xrange(len(self.runs)):
- for j in xrange(len(self.runs[i])):
- # The run name.
- run = self.runs[i][j]
-
- # Function type.
- self.function_type[run] =
relax_data_store.run_types[relax_data_store.run_names.index(run)]
-
- # Store the first non-hybrid run.
- if not self.first_run and self.function_type[run] !=
'hybrid':
- self.first_run = run
+ if type(pipes[0]) == list:
+ # No pipes.
+ if len(pipes[0]) == 0:
+ raise RelaxError, "No pipes are available for use in model
selection in the array " + `pipes[0]` + "."
+
+ # Loop over the data pipes.
+ for i in xrange(len(pipes)):
+ for j in xrange(len(pipes[i])):
+ # Alias the data pipe name.
+ pipe = pipes[i][j]
# Specific duplicate data, number of instances, and model
statistics functions.
- self.duplicate_data[run] =
self.relax.specific_setup.setup('duplicate_data', self.function_type[run])
- self.count_num_instances[run] =
self.relax.specific_setup.setup('num_instances', self.function_type[run])
- self.model_statistics[run] =
self.relax.specific_setup.setup('model_stats', self.function_type[run])
- self.skip_function[run] =
self.relax.specific_setup.setup('skip_function', self.function_type[run])
+ count_num_instances[pipe] = get_specific_fn('num_instances',
relax_data_store[pipe].pipe_type)
+ model_statistics[pipe] = get_specific_fn('model_stats',
relax_data_store[pipe].pipe_type)
+ skip_function[pipe] = get_specific_fn('skip_function',
relax_data_store[pipe].pipe_type)
# Run various tests.
- self.tests(run)
+ tests(pipe)
# All other model selection setup.
else:
- # Loop over the runs.
- for i in xrange(len(self.runs)):
- # The run name.
- run = self.runs[i]
-
- # Function type.
- self.function_type[run] =
relax_data_store.run_types[relax_data_store.run_names.index(run)]
-
- # Store the first non-hybrid run.
- if not self.first_run and self.function_type[run] != 'hybrid':
- self.first_run = run
+ # Loop over the data pipes.
+ for i in xrange(len(pipes)):
+ # Alias the data pipe name.
+ pipe = pipes[i]
# Specific duplicate data, number of instances, and model
statistics functions.
- self.duplicate_data[run] =
self.relax.specific_setup.setup('duplicate_data', self.function_type[run])
- self.count_num_instances[run] =
self.relax.specific_setup.setup('num_instances', self.function_type[run])
- self.model_statistics[run] =
self.relax.specific_setup.setup('model_stats', self.function_type[run])
- self.skip_function[run] =
self.relax.specific_setup.setup('skip_function', self.function_type[run])
+ count_num_instances[pipe] = get_specific_fn('num_instances',
relax_data_store[pipe].pipe_type)
+ model_statistics[pipe] = get_specific_fn('model_stats',
relax_data_store[pipe].pipe_type)
+ skip_function[pipe] = get_specific_fn('skip_function',
relax_data_store[pipe].pipe_type)
# Run various tests.
- self.tests(run)
-
-
- # Number of instances. If the number is not the same for each run, then
the minimum number will give the specific function self.model_statistics the
- # opportunity to consolidate the instances to the minimum number if
possible.
- self.min_instances = 1e99
- self.num_instances = []
- for i in xrange(len(self.runs)):
+ tests(pipe)
+
+
+ # Number of instances. If the number is not the same for each data
pipe, then the minimum
+ # number will give the specific function model_statistics the
opportunity to consolidate the
+ # instances to the minimum number if possible.
+ min_instances = 1e99
+ num_instances = []
+ for i in xrange(len(pipes)):
# An array of arrays - for cross validation model selection.
- if type(self.runs[i]) == list:
- self.num_instances.append([])
+ if type(pipes[i]) == list:
+ num_instances.append([])
# Loop over the nested array.
- for j in xrange(len(self.runs[i])):
+ for j in xrange(len(pipes[i])):
# Number of instances.
- num =
self.count_num_instances[self.runs[i][j]](self.runs[i][j])
- self.num_instances[i].append(num)
+ num = count_num_instances[pipes[i][j]](pipes[i][j])
+ num_instances[i].append(num)
# Minimum.
- if num < self.min_instances:
- self.min_instances = num
+ if num < min_instances:
+ min_instances = num
# All other model selection techniques.
else:
# Number of instances.
- num = self.count_num_instances[self.runs[i]](self.runs[i])
- self.num_instances.append(num)
+ num = count_num_instances[pipes[i]](pipes[i])
+ num_instances.append(num)
# Minimum.
- if num < self.min_instances:
- self.min_instances = num
+ if num < min_instances:
+ min_instances = num
# Loop over the number of instances.
- for i in xrange(self.min_instances):
+ for i in xrange(min_instances):
# Print out.
print "\nInstance " + `i` + ".\n"
print "%-20s %-20s %-20s %-20s %-20s" % ("Run", "Num_params_(k)",
"Num_data_sets_(n)", "Chi2", "Criterion")
@@ -231,24 +214,27 @@
best_model = None
best_crit = 1e300
- # Loop over the runs.
- for j in xrange(len(self.runs)):
+ # Loop over the pipes.
+ for j in xrange(len(pipes)):
# Single-item-out cross validation.
if method == 'CV':
# Sum of chi-squared values.
sum_crit = 0.0
# Loop over the validation samples and sum the chi-squared
values.
- for k in xrange(len(self.runs[j])):
- # Reassign the run.
- run = self.runs[j][k]
+ for k in xrange(len(pipes[j])):
+ # Alias the data pipe name.
+ pipe = pipes[j][k]
+
+ # Switch to this pipe.
+ pipes.switch(pipe)
# Skip function.
- if self.skip_function[run](run=run, instance=i):
+ if skip_function[pipe](instance=i):
continue
# Get the model statistics.
- k, n, chi2 = self.model_statistics[run](run=run,
instance=i, min_instances=self.min_instances)
+ k, n, chi2 = model_statistics[pipe](instance=i,
min_instances=min_instances)
# Missing data sets.
if k == None or n == None or chi2 == None:
@@ -258,47 +244,50 @@
sum_crit = sum_crit + chi2
# Cross-validation criterion (average chi-squared value).
- crit = sum_crit / float(len(self.runs[j]))
+ crit = sum_crit / float(len(pipes[j]))
# Other model selection methods.
else:
- # Reassign the run.
- run = self.runs[j]
+ # Reassign the pipe.
+ pipe = pipes[j]
+
+ # Switch to this pipe.
+ pipes.switch(pipe)
# Skip function.
- if self.skip_function[run](run=run, instance=i,
min_instances=self.min_instances, num_instances=self.num_instances[j]):
+ if skip_function[pipe](instance=i,
min_instances=min_instances, num_instances=num_instances[j]):
continue
# Global stats.
- if self.num_instances[j] > self.min_instances or
self.num_instances[j] == 1:
+ if num_instances[j] > min_instances or num_instances[j] == 1:
global_stats = 1
else:
global_stats = 0
# Get the model statistics.
- k, n, chi2 = self.model_statistics[run](run=run, instance=i,
global_stats=global_stats)
+ k, n, chi2 = model_statistics[pipe](instance=i,
global_stats=global_stats)
# Missing data sets.
if k == None or n == None or chi2 == None:
continue
# Calculate the criterion value.
- crit = self.formula(chi2, float(k), float(n))
+ crit = formula(chi2, float(k), float(n))
# Print out.
- print "%-20s %-20i %-20i %-20.5f %-20.5f" % (run, k, n,
chi2, crit)
+ print "%-20s %-20i %-20i %-20.5f %-20.5f" % (pipe, k, n,
chi2, crit)
# Select model.
if crit < best_crit:
- best_model = run
+ best_model = pipe
best_crit = crit
# Print out of selected model.
- print "\nThe model from the run " + `best_model` + " has been
selected."
-
- # Duplicate the data from the 'best_model' to the model selection
run 'modsel_run'.
+ print "\nThe model from the data pipe " + `best_model` + " has been
selected."
+
+ # Duplicate the data from the 'best_model' to the model selection
data pipe.
if best_model != None:
- self.duplicate_data[best_model](new_run=modsel_run,
old_run=best_model, instance=i, global_stats=global_stats)
+ pipes.copy(best_model, modsel_pipe)
def tests(run):
r5657 - /1.3/generic_fns/model_selection.py