"""
StructureFactorConstraints contains classes for all constraints related experimental static structure factor functions.
.. inheritance-diagram:: fullrmc.Constraints.StructureFactorConstraints
:parts: 1
"""
# standard libraries imports
from __future__ import print_function
import itertools, re
# external libraries imports
import numpy as np
from pdbparser.Utilities.Database import is_element_property, get_element_property
from pdbparser.Utilities.Collection import get_normalized_weighting
# fullrmc imports
from ..Globals import INT_TYPE, FLOAT_TYPE, PI, PRECISION, LOGGER
from ..Globals import str, long, unicode, bytes, basestring, range, xrange, maxint
from ..Core.Collection import is_number, is_integer, get_path
from ..Core.Collection import reset_if_collected_out_of_date, get_real_elements_weight
from ..Core.Collection import get_caller_frames
from ..Core.Constraint import Constraint, ExperimentalConstraint
from ..Core.pairs_histograms import multiple_pairs_histograms_coords, full_pairs_histograms_coords
class StructureFactorConstraint(ExperimentalConstraint):
"""
Controls the Structure Factor noted as S(Q) and also called
total-scattering structure function or Static Structure Factor.
S(Q) is a dimensionless quantity and normalized such as the average
value :math:`<S(Q)>=1`.
It is worth mentioning that S(Q) is nothing other than the normalized and
corrected diffraction pattern if all experimental artefacts powder.
The computation of S(Q) is done through an inverse Sine Fourier transform
of the computed pair distribution function G(r).
.. math::
S(Q) = 1+ \\frac{1}{Q} \\int_{0}^{\\infty} G(r) sin(Qr) dr
From an atomistic model and histogram point of view, G(r) is computed as
the following:
.. math::
G(r) = 4 \\pi r (\\rho_{r} - \\rho_{0})
= 4 \\pi \\rho_{0} r (g(r)-1)
= \\frac{R(r)}{r} - 4 \\pi \\rho_{0}
g(r) is calculated after binning all pair atomic distances into a
weighted histograms as the following:
.. math::
g(r) = \\sum \\limits_{i,j}^{N} w_{i,j} \\frac{\\rho_{i,j}(r)}{\\rho_{0}}
= \\sum \\limits_{i,j}^{N} w_{i,j} \\frac{n_{i,j}(r) / v(r)}{N_{i,j} / V}
Where:\n
:math:`Q` is the momentum transfer. \n
:math:`r` is the distance between two atoms. \n
:math:`\\rho_{i,j}(r)` is the pair density function of atoms i and j. \n
:math:`\\rho_{0}` is the average number density of the system. \n
:math:`w_{i,j}` is the relative weighting of atom types i and j. \n
:math:`R(r)` is the radial distribution function (rdf). \n
:math:`N` is the total number of atoms. \n
:math:`V` is the volume of the system. \n
:math:`n_{i,j}(r)` is the number of atoms i neighbouring j at a distance r. \n
:math:`v(r)` is the annulus volume at distance r and of thickness dr. \n
:math:`N_{i,j}` is the total number of atoms i and j in the system. \n
+----------------------------------------------------------------------+
|.. figure:: reduced_structure_factor_constraint_plot_method.png |
| :width: 530px |
| :height: 400px |
| :align: left |
| |
| Reduced structure factor of memory shape Nickel-Titanium alloy. |
+----------------------------------------------------------------------+
:Parameters:
#. experimentalData (numpy.ndarray, string): Experimental data as
numpy.ndarray or string path to load data using numpy.loadtxt
method.
#. dataWeights (None, numpy.ndarray): Weights array of the same number
of points of experimentalData used in the constraint's standard
error computation. Therefore particular fitting emphasis can be
put on different data points that might be considered as more or less
important in order to get a reasonable and plausible modal.\n
If None is given, all data points are considered of the same
importance in the computation of the constraint's standard error.\n
If numpy.ndarray is given, all weights must be positive and all
zeros weighted data points won't contribute to the total
constraint's standard error. At least a single weight point is
required to be non-zeros and the weights array will be automatically
scaled upon setting such as the the sum of all the weights
is equal to the number of data points.
#. weighting (string): The elements weighting scheme. It must be any
atomic attribute (atomicNumber, neutronCohb, neutronIncohb,
neutronCohXs, neutronIncohXs, atomicWeight, covalentRadius) defined
in pdbparser database. In case of xrays or neutrons experimental
weights, one can simply set weighting to 'xrays' or 'neutrons'
and the value will be automatically adjusted to respectively
'atomicNumber' and 'neutronCohb'. If attribute values are
missing in the pdbparser database, atomic weights must be
given in atomsWeight dictionary argument.
#. atomsWeight (None, dict): Atoms weight dictionary where keys are
atoms element and values are custom weights. If None is given
or partially given, missing elements weighting will be fully set
given weighting scheme.
#. rmin (None, number): The minimum distance value to compute G(r)
histogram. If None is given, rmin is computed as
:math:`2 \\pi / Q_{max}`.
#. rmax (None, number): The maximum distance value to compute G(r)
histogram. If None is given, rmax is computed as
:math:`2 \\pi / dQ`.
#. dr (None, number): The distance bin value to compute G(r)
histogram. If None is given, bin is computed as
:math:`2 \\pi / (Q_{max}-Q_{min})`.
#. scaleFactor (number): A normalization scale factor used to normalize
the computed data to the experimental ones.
#. adjustScaleFactor (list, tuple): Used to adjust fit or guess
the best scale factor during stochastic engine runtime.
It must be a list of exactly three entries.\n
#. The frequency in number of generated moves of finding the best
scale factor. If 0 frequency is given, it means that the scale
factor is fixed.
#. The minimum allowed scale factor value.
#. The maximum allowed scale factor value.
#. windowFunction (None, numpy.ndarray): The window function to
convolute with the computed pair distribution function of the
system prior to comparing it with the experimental data. In
general, the experimental pair distribution function G(r) shows
artificial wrinkles, among others the main reason is because
G(r) is computed by applying a sine Fourier transform to the
experimental structure factor S(q). Therefore window function is
used to best imitate the numerical artefacts in the experimental
data.
#. limits (None, tuple, list): The distance limits to compute the
histograms. If None is given, the limits will be automatically
set the the min and max distance of the experimental data.
Otherwise, a tuple of exactly two items where the first is the
minimum distance or None and the second is the maximum distance
or None.
**NB**: If adjustScaleFactor first item (frequency) is 0, the scale factor
will remain untouched and the limits minimum and maximum won't be checked.
.. code-block:: python
# import fullrmc modules
from fullrmc.Engine import Engine
from fullrmc.Constraints.StructureFactorConstraints import StructureFactorConstraint
# create engine
ENGINE = Engine(path='my_engine.rmc')
# set pdb file
ENGINE.set_pdb('system.pdb')
# create and add constraint
SFC = StructureFactorConstraint(experimentalData="sq.dat", weighting="atomicNumber")
ENGINE.add_constraints(SFC)
"""
def __init__(self, experimentalData, dataWeights=None,
weighting="atomicNumber", atomsWeight=None,
rmin=None, rmax=None, dr=None,
scaleFactor=1.0, adjustScaleFactor=(0, 0.8, 1.2),
windowFunction=None, limits=None):
# initialize variables
self.__experimentalQValues = None
self.__experimentalSF = None
self.__rmin = None
self.__rmax = None
self.__dr = None
self.__minimumDistance = None
self.__maximumDistance = None
self.__bin = None
self.__shellCenters = None
self.__histogramSize = None
self.__shellVolumes = None
self.__Gr2SqMatrix = None
# initialize constraint
super(StructureFactorConstraint, self).__init__( experimentalData=experimentalData, dataWeights=dataWeights, scaleFactor=scaleFactor, adjustScaleFactor=adjustScaleFactor)
# set atomsWeight
self.set_atoms_weight(atomsWeight)
# set elements weighting
self.set_weighting(weighting)
self.__set_weighting_scheme()
# set window function
self.set_window_function(windowFunction)
# set r parameters
self.set_rmin(rmin)
self.set_rmax(rmax)
self.set_dr(dr)
# set frame data
FRAME_DATA = [d for d in self.FRAME_DATA]
FRAME_DATA.extend(['_StructureFactorConstraint__experimentalQValues',
'_StructureFactorConstraint__experimentalSF',
'_StructureFactorConstraint__elementsPairs',
'_StructureFactorConstraint__weightingScheme',
'_StructureFactorConstraint__atomsWeight',
'_StructureFactorConstraint__qmin',
'_StructureFactorConstraint__qmax',
'_StructureFactorConstraint__rmin',
'_StructureFactorConstraint__rmax',
'_StructureFactorConstraint__dr',
'_StructureFactorConstraint__minimumDistance',
'_StructureFactorConstraint__maximumDistance',
'_StructureFactorConstraint__bin',
'_StructureFactorConstraint__shellCenters',
'_StructureFactorConstraint__histogramSize',
'_StructureFactorConstraint__shellVolumes',
'_StructureFactorConstraint__Gr2SqMatrix',
'_StructureFactorConstraint__windowFunction',
'_elementsWeight',] )
RUNTIME_DATA = [d for d in self.RUNTIME_DATA]
RUNTIME_DATA.extend( [] )
object.__setattr__(self, 'FRAME_DATA', tuple(FRAME_DATA) )
object.__setattr__(self, 'RUNTIME_DATA', tuple(RUNTIME_DATA) )
def _codify_update__(self, name='constraint', addDependencies=True):
dependencies = []
code = []
if addDependencies:
code.extend(dependencies)
dw = self.dataWeights
if dw is not None:
dw = list(dw)
code.append("dw = {dw}".format(dw=dw))
wf = self.windowFunction
if isinstance(wf, np.ndarray):
code.append("wf = np.array({wf})".format(wf=list(wf)))
else:
code.append("wf = {wf}".format(wf=wf))
code.append("{name}.set_used({val})".format(name=name, val=self.used))
code.append("{name}.set_scale_factor({val})".format(name=name, val=self.scaleFactor))
code.append("{name}.set_adjust_scale_factor({val})".format(name=name, val=self.adjustScaleFactor))
code.append("{name}.set_data_weights(dw)".format(name=name))
code.append("{name}.set_atoms_weight({val})".format(name=name, val=self.atomsWeight))
code.append("{name}.set_window_function(wf)".format(name=name))
code.append("{name}.set_rmin({val})".format(name=name, val=self.rmin))
code.append("{name}.set_rmax({val})".format(name=name, val=self.rmax))
code.append("{name}.set_dr({val})".format(name=name, val=self.dr))
code.append("{name}.set_limits({val})".format(name=name, val=self.limits))
# return
return dependencies, '\n'.join(code)
def _codify__(self, engine, name='constraint', addDependencies=True):
assert isinstance(name, basestring), LOGGER.error("name must be a string")
assert re.match('[a-zA-Z_][a-zA-Z0-9_]*$', name) is not None, LOGGER.error("given name '%s' can't be used as a variable name"%name)
klass = self.__class__.__name__
dependencies = ['import numpy as np','from fullrmc.Constraints import StructureFactorConstraints']
code = []
if addDependencies:
code.extend(dependencies)
x = list(self.experimentalData[:,0])
y = list(self.experimentalData[:,1])
code.append("x = {x}".format(x=x))
code.append("y = {y}".format(y=y))
code.append("d = np.transpose([x,y]).astype(np.float32)")
dw = self.dataWeights
if dw is not None:
dw = list(dw)
code.append("dw = {dw}".format(dw=dw))
wf = self.windowFunction
if isinstance(wf, np.ndarray):
code.append("wf = np.array({wf})".format(wf=list(wf)))
else:
code.append("wf = {wf}".format(wf=wf))
code.append("{name} = {klass}s.{klass}\
(experimentalData=d, dataWeights=dw, weighting='{weighting}', atomsWeight={atomsWeight}, \
rmin={rmin}, rmax={rmax}, dr={dr}, scaleFactor={scaleFactor}, adjustScaleFactor={adjustScaleFactor}, \
shapeFuncParams=sfp, windowFunction=wf, limits={limits})".format(name=name, klass=klass,
weighting=self.weighting, atomsWeight=self.atomsWeight, rmin=self.rmin,
rmax=self.rmax, dr=self.dr, scaleFactor=self.scaleFactor,
adjustScaleFactor=self.adjustScaleFactor, limits=self.limits))
code.append("{engine}.add_constraints([{name}])".format(engine=engine, name=name))
# return
return dependencies, '\n'.join(code)
#def __getstate__(self):
# # make sure that __Gr2SqMatrix is not pickled but saved to the disk as None
# state = super(StructureFactorConstraint, self).__getstate__()
# state["_StructureFactorConstraint__Gr2SqMatrix"] = None
# return state
#
#def __setstate__(self, state):
# # make sure to regenerate G(r) to S(q) matrix at loading time
# self.__dict__.update( state )
# self.__set_Gr_2_Sq_matrix()
#
def __set_Gr_2_Sq_matrix(self):
if self.__experimentalQValues is None or self.__shellCenters is None:
self.__Gr2SqMatrix = None
else:
Qs = self.__experimentalQValues
Rs = self.__shellCenters
dr = self.__shellCenters[1]-self.__shellCenters[0]
qr = Rs.reshape((-1,1))*(np.ones((len(Rs),1), dtype=FLOAT_TYPE)*Qs)
sinqr = np.sin(qr)
sinqr_q = sinqr/Qs
self.__Gr2SqMatrix = dr*sinqr_q
def __set_weighting_scheme(self):
if self.engine is not None:
self.__elementsPairs = sorted(itertools.combinations_with_replacement(self.engine.elements,2))
#elementsWeight = dict([(el,float(get_element_property(el,self.__weighting))) for el in self.engine.elements])
#self._elementsWeight = dict([(el,self.__atomsWeight.get(el, float(get_element_property(el,self.__weighting)))) for el in self.engine.elements])
self._elementsWeight = get_real_elements_weight(elements=self.engine.elements, weightsDict=self.__atomsWeight, weighting=self.__weighting)
self.__weightingScheme = get_normalized_weighting(numbers=self.engine.numberOfAtomsPerElement, weights=self._elementsWeight)
for k in self.__weightingScheme:
self.__weightingScheme[k] = FLOAT_TYPE(self.__weightingScheme[k])
else:
self.__elementsPairs = None
self.__weightingScheme = None
# dump to repository
self._dump_to_repository({'_StructureFactorConstraint__elementsPairs' : self.__elementsPairs,
'_StructureFactorConstraint__weightingScheme': self.__weightingScheme})
def __set_histogram(self):
if self.__minimumDistance is None or self.__maximumDistance is None or self.__bin is None:
self.__shellCenters = None
self.__histogramSize = None
self.__shellVolumes = None
else:
# compute edges
if self.engine is not None and self.rmax is None:
minHalfBox = np.min( [np.linalg.norm(v)/2. for v in self.engine.basisVectors])
self.__edges = np.arange(self.__minimumDistance,minHalfBox, self.__bin).astype(FLOAT_TYPE)
else:
self.__edges = np.arange(self.__minimumDistance, self.__maximumDistance+self.__bin, self.__bin).astype(FLOAT_TYPE)
# adjust rmin and rmax
self.__minimumDistance = self.__edges[0]
self.__maximumDistance = self.__edges[-1]
# compute shellCenters
self.__shellCenters = (self.__edges[0:-1]+self.__edges[1:])/FLOAT_TYPE(2.)
# set histogram size
self.__histogramSize = INT_TYPE( len(self.__edges)-1 )
# set shell centers and volumes
self.__shellVolumes = FLOAT_TYPE(4.0/3.)*PI*((self.__edges[1:])**3 - self.__edges[0:-1]**3)
# dump to repository
self._dump_to_repository({'_StructureFactorConstraint__minimumDistance': self.__minimumDistance,
'_StructureFactorConstraint__maximumDistance': self.__maximumDistance,
'_StructureFactorConstraint__shellCenters' : self.__shellCenters,
'_StructureFactorConstraint__histogramSize' : self.__histogramSize,
'_StructureFactorConstraint__shellVolumes' : self.__shellVolumes})
# reset constraint
self.reset_constraint()
# reset sq matrix
self.__set_Gr_2_Sq_matrix()
def _on_collector_reset(self):
pass
@property
def rmin(self):
""" Histogram minimum distance. """
return self.__rmin
@property
def rmax(self):
""" Histogram maximum distance. """
return self.__rmax
@property
def dr(self):
""" Histogram bin size."""
return self.__dr
@property
def bin(self):
""" Computed histogram distance bin size."""
return self.__bin
@property
def minimumDistance(self):
""" Computed histogram minimum distance. """
return self.__minimumDistance
@property
def maximumDistance(self):
""" Computed histogram maximum distance. """
return self.__maximumDistance
@property
def qmin(self):
""" Experimental data reciprocal distances minimum. """
return self.__qmin
@property
def qmax(self):
""" Experimental data reciprocal distances maximum. """
return self.__qmax
@property
def dq(self):
""" Experimental data reciprocal distances bin size. """
return self.__experimentalQValues[1]-self.__experimentalQValues[0]
@property
def experimentalQValues(self):
""" Experimental data used q values. """
return self.__experimentalQValues
@property
def histogramSize(self):
""" Histogram size"""
return self.__histogramSize
@property
def shellCenters(self):
""" Shells center array"""
return self.__shellCenters
@property
def shellVolumes(self):
""" Shells volume array"""
return self.__shellVolumes
@property
def experimentalSF(self):
""" Experimental Structure Factor or S(q)"""
return self.__experimentalSF
@property
def elementsPairs(self):
""" Elements pairs """
return self.__elementsPairs
@property
def atomsWeight(self):
"""Custom atoms weight"""
return self.__atomsWeight
@property
def weighting(self):
""" Elements weighting definition. """
return self.__weighting
@property
def weightingScheme(self):
""" Elements weighting scheme. """
return self.__weightingScheme
@property
def windowFunction(self):
""" Convolution window function. """
return self.__windowFunction
@property
def Gr2SqMatrix(self):
""" G(r) to S(q) transformation matrix."""
return self.__Gr2SqMatrix
@property
def _experimentalX(self):
"""For internal use only to interface
ExperimentalConstraint.get_constraints_properties"""
return self.__experimentalQValues
@property
def _experimentalY(self):
"""For internal use only to interface
ExperimentalConstraint.get_constraints_properties"""
return self.__experimentalSF
@property
def _modelX(self):
"""For internal use only to interface
ExperimentalConstraint.get_constraints_properties"""
return self.__experimentalQValues
def listen(self, message, argument=None):
"""
Listens to any message sent from the Broadcaster.
:Parameters:
#. message (object): Any python object to send to constraint's
listen method.
#. argument (object): Any type of argument to pass to the
listeners.
"""
if message in ("engine set","update pdb","update molecules indexes","update elements indexes","update names indexes"):
self.__set_weighting_scheme()
# reset histogram
if self.engine is not None:
self.__set_histogram()
self.reset_constraint() # ADDED 2017-JAN-08
elif message in("update boundary conditions",):
self.reset_constraint()
def set_rmin(self, rmin):
"""
Set rmin value.
:parameters:
#. rmin (None, number): The minimum distance value to compute G(r)
histogram. If None is given, rmin is computed as
:math:`2 \\pi / Q_{max}`.
"""
if rmin is None:
minimumDistance = FLOAT_TYPE( 2.*PI/self.__qmax )
else:
assert is_number(rmin), LOGGER.error("rmin must be None or a number")
minimumDistance = FLOAT_TYPE(rmin)
if self.__maximumDistance is not None:
assert minimumDistance<self.__maximumDistance, LOGGER.error("rmin must be smaller than rmax %s"%self.__maximumDistance)
self.__rmin = rmin
self.__minimumDistance = minimumDistance
# dump to repository
self._dump_to_repository({'_StructureFactorConstraint__rmin': self.__rmin,
'_StructureFactorConstraint__minimumDistance': self.__minimumDistance})
# reset histogram
self.__set_histogram()
def set_rmax(self, rmax):
"""
Set rmax value.
:Parameters:
#. rmax (None, number): The maximum distance value to compute G(r)
histogram. If None is given, rmax is computed as
:math:`2 \\pi / dQ`.
"""
if rmax is None:
dq = self.__experimentalQValues[1]-self.__experimentalQValues[0]
maximumDistance = FLOAT_TYPE( 2.*PI/dq )
else:
assert is_number(rmax), LOGGER.error("rmax must be None or a number")
maximumDistance = FLOAT_TYPE(rmax)
if self.__minimumDistance is not None:
assert maximumDistance>self.__minimumDistance, LOGGER.error("rmax must be bigger than rmin %s"%self.__minimumDistance)
self.__rmax = rmax
self.__maximumDistance = maximumDistance
# dump to repository
self._dump_to_repository({'_StructureFactorConstraint__rmax': self.__rmax,
'_StructureFactorConstraint__maximumDistance': self.__maximumDistance})
# reset histogram
self.__set_histogram()
def set_dr(self, dr):
"""
Set dr value.
:Parameters:
#. dr (None, number): The distance bin value to compute G(r)
histogram. If None is given, bin is computed as
:math:`2 \\pi / (Q_{max}-Q_{min})`.
"""
if dr is None:
bin = 2.*PI/self.__qmax
rbin = round(bin,1)
if rbin>bin:
rbin -= 0.1
bin = FLOAT_TYPE( rbin )
else:
assert is_number(dr), LOGGER.error("dr must be None or a number")
bin = FLOAT_TYPE(dr)
self.__dr = dr
self.__bin = bin
# dump to repository
self._dump_to_repository({'_StructureFactorConstraint__dr': self.__dr,
'_StructureFactorConstraint__bin': self.__bin})
# reset histogram
self.__set_histogram()
def set_weighting(self, weighting):
"""
Set elements weighting. It must be a valid entry of pdbparser atom's
database.
:Parameters:
#. weighting (string): The elements weighting scheme. It must be
any atomic attribute (atomicNumber, neutronCohb, neutronIncohb,
neutronCohXs, neutronIncohXs, atomicWeight, covalentRadius)
defined in pdbparser database. In case of xrays or neutrons
experimental weights, one can simply set weighting to 'xrays'
or 'neutrons' and the value will be automatically adjusted to
respectively 'atomicNumber' and 'neutronCohb'. If attribute
values are missing in the pdbparser database, atomic weights
must be given in atomsWeight dictionary argument.
"""
if weighting.lower() in ["xrays","x-rays","xray","x-ray"]:
LOGGER.fixed("'%s' weighting is set to atomicNumber"%weighting)
weighting = "atomicNumber"
elif weighting.lower() in ["neutron","neutrons"]:
LOGGER.fixed("'%s' weighting is set to neutronCohb"%weighting)
weighting = "neutronCohb"
assert is_element_property(weighting),LOGGER.error( "weighting is not a valid pdbparser atoms database entry")
assert weighting != "atomicFormFactor", LOGGER.error("atomicFormFactor weighting is not allowed")
self.__weighting = weighting
# dump to repository
self._dump_to_repository({'_StructureFactorConstraint__weighting': self.__weighting})
def set_atoms_weight(self, atomsWeight):
"""
Custom set atoms weight. This is the way to setting a atoms weights
different than the given weighting scheme.
:Parameters:
#. atomsWeight (None, dict): Atoms weight dictionary where keys are
atoms element and values are custom weights. If None is given
or partially given, missing elements weighting will be fully set
given weighting scheme.
"""
if atomsWeight is None:
AW = {}
else:
assert isinstance(atomsWeight, dict),LOGGER.error("atomsWeight must be None or a dictionary")
AW = {}
for k in atomsWeight:
assert isinstance(k, basestring),LOGGER.error("atomsWeight keys must be strings")
try:
val = float(atomsWeight[k])
except:
raise LOGGER.error( "atomsWeight values must be numerical")
AW[k]=val
# set atomsWeight
self.__atomsWeight = AW
# dump to repository
self._dump_to_repository({'_StructureFactorConstraint__atomsWeight': self.__atomsWeight})
def set_window_function(self, windowFunction):
"""
Set convolution window function.
:Parameters:
#. windowFunction (None, numpy.ndarray): The window function to
convolute with the computed pair distribution function of the
system prior to comparing it with the experimental data. In
general, the experimental pair distribution function G(r) shows
artificial wrinkles, among others the main reason is because
G(r) is computed by applying a sine Fourier transform to the
experimental structure factor S(q). Therefore window function is
used to best imitate the numerical artefacts in the experimental
data.
"""
if windowFunction is not None:
assert isinstance(windowFunction, np.ndarray), LOGGER.error("windowFunction must be a numpy.ndarray")
assert windowFunction.dtype.type is FLOAT_TYPE, LOGGER.error("windowFunction type must be %s"%FLOAT_TYPE)
assert len(windowFunction.shape) == 1, LOGGER.error("windowFunction must be of dimension 1")
assert len(windowFunction) <= self.experimentalData.shape[0], LOGGER.error("windowFunction length must be smaller than experimental data")
# normalize window function
windowFunction /= np.sum(windowFunction)
# check window size
# set windowFunction
self.__windowFunction = windowFunction
# dump to repository
self._dump_to_repository({'_StructureFactorConstraint__windowFunction': self.__windowFunction})
def set_experimental_data(self, experimentalData):
"""
Set constraint's experimental data.
:Parameters:
#. experimentalData (numpy.ndarray, string): The experimental
data as numpy.ndarray or string path to load data using
numpy.loadtxt function.
"""
# get experimental data
super(StructureFactorConstraint, self).set_experimental_data(experimentalData=experimentalData)
# set limits
self.set_limits(self.limits)
def set_limits(self, limits):
"""
Set the reciprocal distance limits (qmin, qmax).
:Parameters:
#. limits (None, tuple, list): Distance limits to bound
experimental data and compute histograms.
If None is given, the limits will be automatically set to
min and max reciprocal distance recorded in experimental data.
If given, a tuple of minimum reciprocal distance (qmin) or None
and maximum reciprocal distance (qmax) or None should be given.
"""
self._ExperimentalConstraint__set_limits(limits)
# set qvalues
self.__experimentalQValues = self.experimentalData[self.limitsIndexStart:self.limitsIndexEnd+1,0].astype(FLOAT_TYPE)
self.__experimentalSF = self.experimentalData[self.limitsIndexStart:self.limitsIndexEnd+1,1].astype(FLOAT_TYPE)
# set qmin and qmax
self.__qmin = self.__experimentalQValues[0]
self.__qmax = self.__experimentalQValues[-1]
assert self.__qmin>0, LOGGER.error("qmin must be bigger than 0. Experimental null q values are ambigous. Try setting limits.")
# dump to repository
self._dump_to_repository({'_StructureFactorConstraint__experimentalQValues': self.__experimentalQValues,
'_StructureFactorConstraint__experimentalSF' : self.__experimentalSF,
'_StructureFactorConstraint__qmin' : self.__qmin,
'_StructureFactorConstraint__qmax' : self.__qmax})
# set used dataWeights
self._set_used_data_weights(limitsIndexStart=self.limitsIndexStart, limitsIndexEnd=self.limitsIndexEnd)
# reset constraint
self.reset_constraint()
# reset sq matrix
self.__set_Gr_2_Sq_matrix()
def update_standard_error(self):
""" Compute and set constraint's standardError."""
# set standardError
totalSQ = self.get_constraint_value()["total_no_window"]
self.set_standard_error(self.compute_standard_error(modelData = totalSQ))
def check_experimental_data(self, experimentalData):
"""
Check whether experimental data is correct.
:Parameters:
#. experimentalData (object): The experimental data to check.
:Returns:
#. result (boolean): Whether it is correct or not.
#. message (str): Checking message that explains whats's wrong
with the given data
"""
if not isinstance(experimentalData, np.ndarray):
return False, "experimentalData must be a numpy.ndarray"
if experimentalData.dtype.type is not FLOAT_TYPE:
return False, "experimentalData type must be %s"%FLOAT_TYPE
if len(experimentalData.shape) !=2:
return False, "experimentalData must be of dimension 2"
if experimentalData.shape[1] !=2:
return False, "experimentalData must have only 2 columns"
# check distances order
inOrder = (np.array(sorted(experimentalData[:,0]), dtype=FLOAT_TYPE)-experimentalData[:,0])<=PRECISION
if not np.all(inOrder):
return False, "experimentalData distances are not sorted in order"
if experimentalData[0][0]<0:
return False, "experimentalData distances min value is found negative"
# data format is correct
return True, ""
def compute_standard_error(self, modelData):
"""
Compute the standard error (StdErr) as the squared deviations
between model computed data and the experimental ones.
.. math::
StdErr = \\sum \\limits_{i}^{N} W_{i}(Y(X_{i})-F(X_{i}))^{2}
Where:\n
:math:`N` is the total number of experimental data points. \n
:math:`W_{i}` is the data point weight. It becomes equivalent to 1 when dataWeights is set to None. \n
:math:`Y(X_{i})` is the experimental data point :math:`X_{i}`. \n
:math:`F(X_{i})` is the computed from the model data :math:`X_{i}`. \n
:Parameters:
#. modelData (numpy.ndarray): The data to compare with the
experimental one and compute the squared deviation.
:Returns:
#. standardError (number): The calculated constraint's
standardError.
"""
# compute difference
diff = self.__experimentalSF-modelData
# return standard error
if self._usedDataWeights is None:
return np.add.reduce((diff)**2)
else:
return np.add.reduce(self._usedDataWeights*((diff)**2))
def _get_Sq_from_Gr(self, Gr):
return np.sum(Gr.reshape((-1,1))*self.__Gr2SqMatrix, axis=0)+1
def _apply_scale_factor(self, Sq, scaleFactor):
if scaleFactor != 1:
Sq = scaleFactor*(Sq-1) + 1
return Sq
def __get_total_Sq(self, data, rho0):
"""This method is created just to speed up the computation of
the total Sq upon fitting."""
Gr = np.zeros(self.__histogramSize, dtype=FLOAT_TYPE)
for pair in self.__elementsPairs:
# get weighting scheme
wij = self.__weightingScheme.get(pair[0]+"-"+pair[1], None)
if wij is None:
wij = self.__weightingScheme[pair[1]+"-"+pair[0]]
# get number of atoms per element
ni = self.engine.numberOfAtomsPerElement[pair[0]]
nj = self.engine.numberOfAtomsPerElement[pair[1]]
# get index of element
idi = self.engine.elements.index(pair[0])
idj = self.engine.elements.index(pair[1])
# get Nij
if idi == idj:
Nij = ni*(ni-1)/2.0
Dij = Nij/self.engine.volume
nij = data["intra"][idi,idj,:]+data["inter"][idi,idj,:]
Gr += wij*nij/Dij
else:
Nij = ni*nj
Dij = Nij/self.engine.volume
nij = data["intra"][idi,idj,:]+data["intra"][idj,idi,:] + data["inter"][idi,idj,:]+data["inter"][idj,idi,:]
Gr += wij*nij/Dij
# Devide by shells volume
Gr /= self.shellVolumes
# compute total G(r)
#rho0 = (self.engine.numberOfAtoms/self.engine.volume).astype(FLOAT_TYPE)
Gr = (FLOAT_TYPE(4.)*PI*self.__shellCenters*rho0)*(Gr-1)
# Compute S(q) from G(r)
Sq = self._get_Sq_from_Gr(Gr)
# Multiply by scale factor
self._fittedScaleFactor = self.get_adjusted_scale_factor(self.__experimentalSF, Sq, self._usedDataWeights)
# apply scale factor
Sq = self._apply_scale_factor(Sq, self._fittedScaleFactor)
# apply multiframe prior and weight
Sq = self._apply_multiframe_prior(Sq)
# convolve total with window function
if self.__windowFunction is not None:
Sq = np.convolve(Sq, self.__windowFunction, 'same')
return Sq
def get_adjusted_scale_factor(self, experimentalData, modelData, dataWeights):
"""Overload to reduce S(q) prior to fitting scale factor.
S(q) -> 1 at high q and this will create a wrong scale factor.
Overloading can be avoided but it's better to for performance reasons
"""
SF = self.scaleFactor
# check to update scaleFactor
if self.adjustScaleFactorFrequency:
if not self.engine.accepted%self.adjustScaleFactorFrequency:
SF = self.fit_scale_factor(experimentalData-1, modelData-1, dataWeights)
return SF
def _get_constraint_value(self, data, applyMultiframePrior=True):
# http://erice2011.docking.org/upload/Other/Billinge_PDF/03-ReadingMaterial/BillingePDF2011.pdf page 6
#import time
#startTime = time.clock()
output = {}
for pair in self.__elementsPairs:
output["sf_intra_%s-%s" % pair] = np.zeros(self.__histogramSize, dtype=FLOAT_TYPE)
output["sf_inter_%s-%s" % pair] = np.zeros(self.__histogramSize, dtype=FLOAT_TYPE)
output["sf_total_%s-%s" % pair] = np.zeros(self.__histogramSize, dtype=FLOAT_TYPE)
gr = np.zeros(self.__histogramSize, dtype=FLOAT_TYPE)
for pair in self.__elementsPairs:
# get weighting scheme
wij = self.__weightingScheme.get(pair[0]+"-"+pair[1], None)
if wij is None:
wij = self.__weightingScheme[pair[1]+"-"+pair[0]]
# get number of atoms per element
ni = self.engine.numberOfAtomsPerElement[pair[0]]
nj = self.engine.numberOfAtomsPerElement[pair[1]]
# get index of element
idi = self.engine.elements.index(pair[0])
idj = self.engine.elements.index(pair[1])
# get Nij
if idi == idj:
Nij = ni*(ni-1)/2.0
output["sf_intra_%s-%s" % pair] += data["intra"][idi,idj,:]
output["sf_inter_%s-%s" % pair] += data["inter"][idi,idj,:]
else:
Nij = ni*nj
output["sf_intra_%s-%s" % pair] += data["intra"][idi,idj,:] + data["intra"][idj,idi,:]
output["sf_inter_%s-%s" % pair] += data["inter"][idi,idj,:] + data["inter"][idj,idi,:]
# compute g(r)
nij = output["sf_intra_%s-%s" % pair] + output["sf_inter_%s-%s" % pair]
dij = nij/self.__shellVolumes
Dij = Nij/self.engine.volume
gr += wij*dij/Dij
# calculate intensityFactor
intensityFactor = (self.engine.volume*wij)/(Nij*self.__shellVolumes)
# divide by factor
output["sf_intra_%s-%s" % pair] *= intensityFactor
output["sf_inter_%s-%s" % pair] *= intensityFactor
output["sf_total_%s-%s" % pair] = output["sf_intra_%s-%s" % pair] + output["sf_inter_%s-%s" % pair]
# Compute S(q) from G(r)
output["sf_intra_%s-%s" % pair] = self._get_Sq_from_Gr(output["sf_intra_%s-%s" % pair])
output["sf_inter_%s-%s" % pair] = self._get_Sq_from_Gr(output["sf_inter_%s-%s" % pair])
output["sf_total_%s-%s" % pair] = self._get_Sq_from_Gr(output["sf_total_%s-%s" % pair])
# compute total G(r)
rho0 = (self.engine.numberOfAtoms/self.engine.volume).astype(FLOAT_TYPE)
Gr = (FLOAT_TYPE(4.)*PI*self.__shellCenters*rho0) * (gr-1)
# Compute S(q) from G(r)
Sq = self._get_Sq_from_Gr(Gr)
# multiply by scale factor
output["total_no_window"] = self._apply_scale_factor(Sq, self._fittedScaleFactor)
# apply multiframe prior and weight
if applyMultiframePrior:
output["total_no_window"] = self._apply_multiframe_prior(output["total_no_window"])
# convolve total with window function
if self.__windowFunction is not None:
output["total"] = np.convolve(output["total_no_window"], self.__windowFunction, 'same').astype(FLOAT_TYPE)
else:
output["total"] = output["total_no_window"]
return output
def get_constraint_value(self, applyMultiframePrior=True):
"""
Compute all partial Structure Factor (SQs).
:Parameters:
#. applyMultiframePrior (boolean): Whether to apply subframe weight
and prior to the total. This will only have an effect when used
frame is a subframe and in case subframe weight and prior is
defined.
:Returns:
#. SQs (dictionary): The SQs dictionnary, where keys are the
element wise intra and inter molecular SQs and values are
the computed SQs.
"""
if self.data is None:
LOGGER.warn("data must be computed first using 'compute_data' method.")
return {}
return self._get_constraint_value(self.data, applyMultiframePrior=applyMultiframePrior)
def get_constraint_original_value(self):
"""
Compute all partial Pair Distribution Functions (PDFs).
:Returns:
#. PDFs (dictionary): The PDFs dictionnary, where keys are the
element wise intra and inter molecular PDFs and values are the
computed PDFs.
"""
if self.originalData is None:
LOGGER.warn("originalData must be computed first using 'compute_data' method.")
return {}
return self._get_constraint_value(self.originalData)
@reset_if_collected_out_of_date
def compute_data(self, update=True):
""" Compute constraint's data.
:Parameters:
#. update (boolean): whether to update constraint data and
standard error with new computation. If data is computed and
updated by another thread or process while the stochastic
engine is running, this might lead to a state alteration of
the constraint which will lead to a no additional accepted
moves in the run
:Returns:
#. data (dict): constraint data dictionary
#. standardError (float): constraint standard error
"""
intra,inter = full_pairs_histograms_coords( boxCoords = self.engine.boxCoordinates,
basis = self.engine.basisVectors,
isPBC = self.engine.isPBC,
moleculeIndex = self.engine.moleculesIndex,
elementIndex = self.engine.elementsIndex,
numberOfElements = self.engine.numberOfElements,
minDistance = self.__minimumDistance,
maxDistance = self.__maximumDistance,
histSize = self.__histogramSize,
bin = self.__bin,
ncores = self.engine._runtime_ncores )
# create data and compute standard error
data = {"intra":intra, "inter":inter}
totalSQ = self.__get_total_Sq(data, rho0=self.engine.numberDensity)
stdError = self.compute_standard_error(modelData = totalSQ)
# update
if update:
self.set_data(data)
self.set_active_atoms_data_before_move(None)
self.set_active_atoms_data_after_move(None)
self.set_standard_error(stdError)
# set original data
if self.originalData is None:
self._set_original_data(self.data)
# return
return data, stdError
def compute_before_move(self, realIndexes, relativeIndexes):
"""
Compute constraint before move is executed
:Parameters:
#. realIndexes (numpy.ndarray): Not used here.
#. relativeIndexes (numpy.ndarray): Group atoms relative index
the move will be applied to.
"""
intraM,interM = multiple_pairs_histograms_coords( indexes = relativeIndexes,
boxCoords = self.engine.boxCoordinates,
basis = self.engine.basisVectors,
isPBC = self.engine.isPBC,
moleculeIndex = self.engine.moleculesIndex,
elementIndex = self.engine.elementsIndex,
numberOfElements = self.engine.numberOfElements,
minDistance = self.__minimumDistance,
maxDistance = self.__maximumDistance,
histSize = self.__histogramSize,
bin = self.__bin,
allAtoms = True,
ncores = self.engine._runtime_ncores )
intraF,interF = full_pairs_histograms_coords( boxCoords = self.engine.boxCoordinates[relativeIndexes],
basis = self.engine.basisVectors,
isPBC = self.engine.isPBC,
moleculeIndex = self.engine.moleculesIndex[relativeIndexes],
elementIndex = self.engine.elementsIndex[relativeIndexes],
numberOfElements = self.engine.numberOfElements,
minDistance = self.__minimumDistance,
maxDistance = self.__maximumDistance,
histSize = self.__histogramSize,
bin = self.__bin,
ncores = self.engine._runtime_ncores )
self.set_active_atoms_data_before_move( {"intra":intraM-intraF, "inter":interM-interF} )
self.set_active_atoms_data_after_move(None)
def compute_after_move(self, realIndexes, relativeIndexes, movedBoxCoordinates):
"""
Compute constraint after move is executed
:Parameters:
#. realIndexes (numpy.ndarray): Not used here.
#. relativeIndexes (numpy.ndarray): Group atoms relative index
the move will be applied to.
#. movedBoxCoordinates (numpy.ndarray): The moved atoms new coordinates.
"""
# change coordinates temporarily
boxData = np.array(self.engine.boxCoordinates[relativeIndexes], dtype=FLOAT_TYPE)
self.engine.boxCoordinates[relativeIndexes] = movedBoxCoordinates
# calculate pair distribution function
intraM,interM = multiple_pairs_histograms_coords( indexes = relativeIndexes,
boxCoords = self.engine.boxCoordinates,
basis = self.engine.basisVectors,
isPBC = self.engine.isPBC,
moleculeIndex = self.engine.moleculesIndex,
elementIndex = self.engine.elementsIndex,
numberOfElements = self.engine.numberOfElements,
minDistance = self.__minimumDistance,
maxDistance = self.__maximumDistance,
histSize = self.__histogramSize,
bin = self.__bin,
allAtoms = True,
ncores = self.engine._runtime_ncores )
intraF,interF = full_pairs_histograms_coords( boxCoords = self.engine.boxCoordinates[relativeIndexes],
basis = self.engine.basisVectors,
isPBC = self.engine.isPBC,
moleculeIndex = self.engine.moleculesIndex[relativeIndexes],
elementIndex = self.engine.elementsIndex[relativeIndexes],
numberOfElements = self.engine.numberOfElements,
minDistance = self.__minimumDistance,
maxDistance = self.__maximumDistance,
histSize = self.__histogramSize,
bin = self.__bin,
ncores = self.engine._runtime_ncores )
# set active atoms data
self.set_active_atoms_data_after_move( {"intra":intraM-intraF, "inter":interM-interF} )
# reset coordinates
self.engine.boxCoordinates[relativeIndexes] = boxData
# compute standardError after move
dataIntra = self.data["intra"]-self.activeAtomsDataBeforeMove["intra"]+self.activeAtomsDataAfterMove["intra"]
dataInter = self.data["inter"]-self.activeAtomsDataBeforeMove["inter"]+self.activeAtomsDataAfterMove["inter"]
totalSQ = self.__get_total_Sq({"intra":dataIntra, "inter":dataInter}, rho0=self.engine.numberDensity)
self.set_after_move_standard_error( self.compute_standard_error(modelData = totalSQ) )
# increment tried
self.increment_tried()
def accept_move(self, realIndexes, relativeIndexes):
"""
Accept move
:Parameters:
#. realIndexes (numpy.ndarray): Not used here.
#. relativeIndexes (numpy.ndarray): Not used here.
"""
dataIntra = self.data["intra"]-self.activeAtomsDataBeforeMove["intra"]+self.activeAtomsDataAfterMove["intra"]
dataInter = self.data["inter"]-self.activeAtomsDataBeforeMove["inter"]+self.activeAtomsDataAfterMove["inter"]
# change permanently _data
self.set_data( {"intra":dataIntra, "inter":dataInter} )
# reset activeAtoms data
self.set_active_atoms_data_before_move(None)
self.set_active_atoms_data_after_move(None)
# update standardError
self.set_standard_error( self.afterMoveStandardError )
self.set_after_move_standard_error( None )
# set new scale factor
self._set_fitted_scale_factor_value(self._fittedScaleFactor)
# increment accepted
self.increment_accepted()
def reject_move(self, realIndexes, relativeIndexes):
"""
Reject move
:Parameters:
#. realIndexes (numpy.ndarray): Not used here.
#. relativeIndexes (numpy.ndarray): Not used here.
"""
# reset activeAtoms data
self.set_active_atoms_data_before_move(None)
self.set_active_atoms_data_after_move(None)
# update standardError
self.set_after_move_standard_error( None )
def compute_as_if_amputated(self, realIndex, relativeIndex):
"""
Compute and return constraint's data and standard error as if
given atom is amputated.
:Parameters:
#. realIndex (numpy.ndarray): Atom's index as a numpy array
of a single element.
#. relativeIndex (numpy.ndarray): Atom's relative index as a
numpy array of a single element.
"""
# compute data
self.compute_before_move(realIndexes=realIndex, relativeIndexes=relativeIndex)
dataIntra = self.data["intra"]-self.activeAtomsDataBeforeMove["intra"]
dataInter = self.data["inter"]-self.activeAtomsDataBeforeMove["inter"]
data = {"intra":dataIntra, "inter":dataInter}
# temporarily adjust self.__weightingScheme
weightingScheme = self.__weightingScheme
relativeIndex = relativeIndex[0]
selectedElement = self.engine.allElements[relativeIndex]
self.engine.numberOfAtomsPerElement[selectedElement] -= 1
self.__weightingScheme = get_normalized_weighting(numbers=self.engine.numberOfAtomsPerElement, weights=self._elementsWeight )
for k in self.__weightingScheme:
self.__weightingScheme[k] = FLOAT_TYPE(self.__weightingScheme[k])
## END OF ADDED 08 FEB 2017
# compute standard error
if not self.engine._RT_moveGenerator.allowFittingScaleFactor:
SF = self.adjustScaleFactorFrequency
self._set_adjust_scale_factor_frequency(0)
rho0 = ((self.engine.numberOfAtoms-1)/self.engine.volume).astype(FLOAT_TYPE)
totalSQ = self.__get_total_Sq(data, rho0=rho0)
standardError = self.compute_standard_error(modelData = totalSQ)
if not self.engine._RT_moveGenerator.allowFittingScaleFactor:
self._set_adjust_scale_factor_frequency(SF)
# reset activeAtoms data
self.set_active_atoms_data_before_move(None)
# set amputation
self.set_amputation_data( {'data':data, 'weightingScheme':self.__weightingScheme} )
# compute standard error
self.set_amputation_standard_error( standardError )
# reset weightingScheme and number of atoms per element
self.__weightingScheme = weightingScheme
self.engine.numberOfAtomsPerElement[selectedElement] += 1
def accept_amputation(self, realIndex, relativeIndex):
"""
Accept amputated atom and sets constraints data and standard error accordingly.
:Parameters:
#. realIndex (numpy.ndarray): Not used here.
#. relativeIndex (numpy.ndarray): Not used here.
"""
#self.set_data( self.amputationData ) ## COMMENTED 08 FEB 2017
self.set_data( self.amputationData['data'] )
self.__weightingScheme = self.amputationData['weightingScheme']
self.set_standard_error( self.amputationStandardError )
self.set_amputation_data( None )
self.set_amputation_standard_error( None )
# set new scale factor
self._set_fitted_scale_factor_value(self._fittedScaleFactor)
def reject_amputation(self, realIndex, relativeIndex):
"""
Reject amputated atom and set constraint's data and standard
error accordingly.
:Parameters:
#. realIndex (numpy.ndarray): Not used here.
#. relativeIndex (numpy.ndarray): Not used here.
"""
self.set_amputation_data( None )
self.set_amputation_standard_error( None )
def _on_collector_collect_atom(self, realIndex):
pass
def _on_collector_release_atom(self, realIndex):
pass
def _constraint_copy_needs_lut(self):
return {'_StructureFactorConstraint__elementsPairs' :'_StructureFactorConstraint__elementsPairs',
'_StructureFactorConstraint__histogramSize' :'_StructureFactorConstraint__histogramSize',
'_StructureFactorConstraint__weightingScheme' :'_StructureFactorConstraint__weightingScheme',
'_StructureFactorConstraint__shellVolumes' :'_StructureFactorConstraint__shellVolumes',
'_StructureFactorConstraint__shellCenters' :'_StructureFactorConstraint__shellCenters',
'_StructureFactorConstraint__windowFunction' :'_StructureFactorConstraint__windowFunction',
'_StructureFactorConstraint__experimentalQValues' :'_StructureFactorConstraint__experimentalQValues',
'_StructureFactorConstraint__experimentalSF' :'_StructureFactorConstraint__experimentalSF',
'_StructureFactorConstraint__Gr2SqMatrix' :'_StructureFactorConstraint__Gr2SqMatrix',
'_StructureFactorConstraint__minimumDistance' :'_StructureFactorConstraint__minimumDistance',
'_StructureFactorConstraint__maximumDistance' :'_StructureFactorConstraint__maximumDistance',
'_StructureFactorConstraint__bin' :'_StructureFactorConstraint__bin',
'_ExperimentalConstraint__scaleFactor' :'_ExperimentalConstraint__scaleFactor',
'_ExperimentalConstraint__dataWeights' :'_ExperimentalConstraint__dataWeights',
'_ExperimentalConstraint__multiframePrior' :'_ExperimentalConstraint__multiframePrior',
'_ExperimentalConstraint__multiframeWeight' :'_ExperimentalConstraint__multiframeWeight',
'_ExperimentalConstraint__limits' :'_ExperimentalConstraint__limits',
'_ExperimentalConstraint__limitsIndexStart' :'_ExperimentalConstraint__limitsIndexStart',
'_ExperimentalConstraint__limitsIndexEnd' :'_ExperimentalConstraint__limitsIndexEnd',
'_Constraint__used' :'_Constraint__used',
'_Constraint__data' :'_Constraint__data',
'_Constraint__state' :'_Constraint__state',
'_Constraint__standardError' :'_Constraint__standardError',
'_fittedScaleFactor' :'_fittedScaleFactor',
'_usedDataWeights' :'_usedDataWeights',
'_Engine__state' :'_Engine__state',
'_Engine__boxCoordinates' :'_Engine__boxCoordinates',
'_Engine__basisVectors' :'_Engine__basisVectors',
'_Engine__isPBC' :'_Engine__isPBC',
'_Engine__moleculesIndex' :'_Engine__moleculesIndex',
'_Engine__elementsIndex' :'_Engine__elementsIndex',
'_Engine__numberOfAtomsPerElement' :'_Engine__numberOfAtomsPerElement',
'_Engine__elements' :'_Engine__elements',
'_Engine__numberDensity' :'_Engine__numberDensity',
'_Engine__volume' :'_Engine__volume',
'_Engine__realCoordinates' :'_Engine__realCoordinates',
'_atomsCollector' :'_atomsCollector',
('engine','_atomsCollector') :'_atomsCollector',
}
def plot(self, xlabelParams={'xlabel':'$Q(\\AA^{-1})$', 'size':10},
ylabelParams={'ylabel':'$S(Q)$', 'size':10},
**kwargs):
"""
Alias to ExperimentalConstraint.plot with additional parameters
:Additional/Adjusted Parameters:
#. xlabelParams (None, dict): modified matplotlib.axes.Axes.set_xlabel
parameters.
#. ylabelParams (None, dict): modified matplotlib.axes.Axes.set_ylabel
parameters.
"""
return super(StructureFactorConstraint, self).plot(xlabelParams= xlabelParams,
ylabelParams= ylabelParams,
**kwargs)
class ReducedStructureFactorConstraint(StructureFactorConstraint):
"""
The Reduced Structure Factor that we will also note S(Q)
is exactly the same quantity as the Structure Factor but with
the slight difference that it is normalized to 0 rather than 1
and therefore :math:`<S(Q)>=0`.
The computation of S(Q) is done through a Sine inverse Fourier transform
of the computed pair distribution function noted as G(r).
.. math::
S(Q) = \\frac{1}{Q} \\int_{0}^{\\infty} G(r) sin(Qr) dr
The only reason why the Reduced Structure Factor is implemented, is because
many experimental data are treated in this form. And it is just convenient
not to manipulate the experimental data every time.
"""
def _get_Sq_from_Gr(self, Gr):
return np.sum(Gr.reshape((-1,1))*self.Gr2SqMatrix, axis=0)
def _apply_scale_factor(self, Sq, scaleFactor):
if scaleFactor != 1:
Sq = scaleFactor*Sq
return Sq
def get_adjusted_scale_factor(self, experimentalData, modelData, dataWeights):
""" dummy overload that does exactly the same thing
"""
SF = self.scaleFactor
# check to update scaleFactor
if self.adjustScaleFactorFrequency:
if not self.engine.accepted%self.adjustScaleFactorFrequency:
SF = self.fit_scale_factor(experimentalData, modelData, dataWeights)
return SF
def plot(self, xlabelParams={'xlabel':'$Q(\\AA^{-1})$', 'size':10},
ylabelParams={'ylabel':'$S(Q)-1$', 'size':10},
**kwargs):
"""
Alias to ExperimentalConstraint.plot with additional parameters
:Additional/Adjusted Parameters:
#. xlabelParams (None, dict): modified matplotlib.axes.Axes.set_xlabel
parameters.
#. ylabelParams (None, dict): modified matplotlib.axes.Axes.set_ylabel
parameters.
"""
return super(StructureFactorConstraint, self).plot(xlabelParams= xlabelParams,
ylabelParams= ylabelParams,
**kwargs)
