infrared.infrared.Model Class Reference

Public Member Functions
	__init__ (self, number=None, domain=None, name='X')
	Init model.
	add_variables (self, number, domain, name='X')
	Add variable domains.
	restrict_domains (self, vars, domain)
	Restrict the domain of a variable.
	add_constraints (self, constraints)
	Add constraints to the model.
	add_functions (self, functions, group='base')
	Add functions to the model.
	num_named_variables (self, name)
	Number of variables.
	has_empty_domains (self)
	Check inconsistency due to empty domains.
	add_feature (self, name, group, eval_fun)
	Add a (custom) feature.
	eval_feature (self, assignment, name)
	Evaluate named feature at assignment.
	set_feature_weight (self, weight, name)
	Set the weight of a feature and its function group.
	idx (self, variables)
	Raw indices of named variables.
	dependencies (self, non_redundant=True)
	Dependencies due to constraints and functions.
	bindependencies (self, non_redundant=True)
	write_graph (self, out, non_redundant=True)
	Write dependency graph.
	connected_components (self)
	Connected components of the model's dependency graph.

Protected Member Functions
	_automatic_feature (self, group)
	Automatic feature for function group.

Static Protected Member Functions
	_expand_to_cliques (dependencies)
	Expand non-binary dependencies to cliques of binary deps.
	_remove_subsumed_dependencies (deps)
	Removes redundant dependencies that are subsumed by others Removes all dependencies that are subsumed by other dependencies in deps.

Protected Attributes
	_constraints
	_functions
	_domains
	_features

Properties
	num_variables = property
	Number of all variables.
	constraints = property
	Constraints of the model.
	functions = property
	All functions of the model.
	domains = property
	Domains of the model.
	features = property
	Features of the model.

Constructor & Destructor Documentation

__init__()

infrared.infrared.Model.__init__	(		self,
			number = None,
			domain = None,
			name = 'X'
	)

Init model.

   If number is not None, calls add_variables with the
   given parameters after initialization.

Parameters

number	if not None, number of variables
domain	domains of variables
name	name of variable series. Defaults to X

Member Function Documentation

_automatic_feature()

infrared.infrared.Model._automatic_feature (   self,   group  )

protected

Automatic feature for function group.

Feature with value that is derived as sum of the feature functions

Returns

the feature

_expand_to_cliques()

infrared.infrared.Model._expand_to_cliques (   dependencies )

staticprotected

Expand non-binary dependencies to cliques of binary deps.

Parameters

dependencies

list of dependencies

Returns

list of binary dependencies

_remove_subsumed_dependencies()

infrared.infrared.Model._remove_subsumed_dependencies (   deps )

staticprotected

Removes redundant dependencies that are subsumed by others Removes all dependencies that are subsumed by other dependencies in deps.

Parameters

deps	list of dependencies (where a dependency is a list of indices)

Returns

pruned list of dependencies

add_constraints()

infrared.infrared.Model.add_constraints	(		self,
			constraints
	)

Add constraints to the model.

Parameters

constraints

an iterable of constraints or a single constraint

Note

supports optimizations via on_add and entailed methods of added constraints; see ValueIn

add_feature()

infrared.infrared.Model.add_feature	(		self,
			name,
			group,
			eval_fun
	)

Add a (custom) feature.

Parameters

name	name of the feature
group	one or several groups of feature controlled functions
eval_fun	function to evaluate the feature at an assignment

add_functions()

infrared.infrared.Model.add_functions	(		self,
			functions,
			group = 'base'
	)

Add functions to the model.

Parameters

functions	[const] an iterable of functions or a single function
group	indentifier of function group. Defaults to base

Note

deep copies the input functions

add_variables()

infrared.infrared.Model.add_variables	(		self,
			number,
			domain,
			name = 'X'
	)

Add variable domains.

Parameters

number	number of variables
domain	domain size; defines the domain values as 0..domain-1
name	assign a name to the variable(s)

bindependencies()

infrared.infrared.Model.bindependencies	(		self,
			non_redundant = True
	)

connected_components()

infrared.infrared.Model.connected_components

(

self

)

Connected components of the model's dependency graph.

Returns

a list of sets of the connected components

dependencies()

infrared.infrared.Model.dependencies	(		self,
			non_redundant = True
	)

Dependencies due to constraints and functions.

Parameters

non_redundant

whether the dependency list is made non-redundant. Defaults to True

Returns

list of lists of indices of variables that depend on each other either through functions or constraints

eval_feature()

infrared.infrared.Model.eval_feature	(		self,
			assignment,
			name
	)

Evaluate named feature at assignment.

Parameters

assignment	the assignment
name	name of the feature

Returns

value of the feature

has_empty_domains()

infrared.infrared.Model.has_empty_domains

(

self

)

Check inconsistency due to empty domains.

Returns

whether model has empty domains

idx()

infrared.infrared.Model.idx	(		self,
			variables
	)

Raw indices of named variables.

Parameters

variables

single variable or list of variables; variables can be named; default name 'X'

Returns

list of (internal) variable indices

num_named_variables()

infrared.infrared.Model.num_named_variables	(		self,
			name
	)

Number of variables.

Parameters

name	name of the variables series

Returns

number of variables of the given series

restrict_domains()

infrared.infrared.Model.restrict_domains	(		self,
			vars,
			domain
	)

Restrict the domain of a variable.

Parameters

vars	variable or list of variables, each specified by (name,index); or simply index, then addressing ('X',index)
domain	the domain

Note

the domain bounds are intersected with the original domain

set_feature_weight()

infrared.infrared.Model.set_feature_weight	(		self,
			weight,
			name
	)

Set the weight of a feature and its function group.

   This method sets the weight for the Feature itself and in
   all the functions (currently!) in its group.

   The method should be called after all functions of its group are
   defined. Otherwise, weights of the feature and its functions get out
   of sync (but can be re-synced by another call to this method.)

Parameters

weight	the weight
name	the feature name

write_graph()

infrared.infrared.Model.write_graph	(		self,
			out,
			non_redundant = True
	)

Write dependency graph.

Writes the dependency graph to file in dot format; hyper-edges are expanded to cliques.

Parameters

out	a filehandle of name of the target file

Member Data Documentation

_constraints

infrared.infrared.Model._constraints

protected

_domains

infrared.infrared.Model._domains

protected

_features

infrared.infrared.Model._features

protected

_functions

infrared.infrared.Model._functions

protected

Property Documentation

constraints

infrared.infrared.Model.constraints = property

static

Constraints of the model.

Returns

specification of the model's functions

domains

infrared.infrared.Model.domains = property

static

Domains of the model.

Returns

list of all domain descriptions

features

infrared.infrared.Model.features = property

static

Features of the model.

Returns

dictionary of features

functions

infrared.infrared.Model.functions = property

static

All functions of the model.

Returns

list of all functions

num_variables

infrared.infrared.Model.num_variables = property

static

Number of all variables.

Returns

number of all variables

---

The documentation for this class was generated from the following file:

• src/infrared/infrared.py