Source code for schrodinger.application.matsci.gsas.GSASIIpwd

#/usr/bin/env python
# -*- coding: utf-8 -*-
'''
# Third-party code. No Schrodinger Copyright.

*GSASII powder calculation module*
==================================

'''
########### SVN repository information ###################
# $Date: 2019-10-01 16:14:03 -0400 (Tue, 01 Oct 2019) $
# $Author: vondreele $
# $Revision: 4166 $
# $URL: https://subversion.xray.aps.anl.gov/pyGSAS/trunk/GSASIIpwd.py $
# $Id: GSASIIpwd.py 4166 2019-10-01 20:14:03Z vondreele $
########### SVN repository information ###################
# flake8: noqa

import copy
import math
import os
import random as rand
import subprocess as subp
import sys
import time

import numpy as np
import numpy.fft as fft
import numpy.linalg as nl
import numpy.ma as ma
import scipy.interpolate as si
import scipy.optimize as so
import scipy.special as sp
import scipy.stats as st

from . import GSASIIElem as G2elem
from . import GSASIIfiles as G2fil
from . import GSASIIlattice as G2lat
from . import GSASIImath as G2mth
from . import GSASIIspc as G2spc

try:
    import pypowder as pyd
except ImportError:
    pass
try:
    import pydiffax as pyx
except ImportError:
    pass

# trig functions in degrees
tand = lambda x: math.tan(x * math.pi / 180.)
atand = lambda x: 180. * math.atan(x) / math.pi
atan2d = lambda y, x: 180. * math.atan2(y, x) / math.pi
cosd = lambda x: math.cos(x * math.pi / 180.)
acosd = lambda x: 180. * math.acos(x) / math.pi
rdsq2d = lambda x, p: round(1.0 / math.sqrt(x), p)
#numpy versions
npsind = lambda x: np.sin(x * np.pi / 180.)
npasind = lambda x: 180. * np.arcsin(x) / math.pi
npcosd = lambda x: np.cos(x * math.pi / 180.)
npacosd = lambda x: 180. * np.arccos(x) / math.pi
nptand = lambda x: np.tan(x * math.pi / 180.)
npatand = lambda x: 180. * np.arctan(x) / np.pi
npatan2d = lambda y, x: 180. * np.arctan2(y, x) / np.pi
npT2stl = lambda tth, wave: 2.0 * npsind(tth / 2.0) / wave  #=d*
npT2q = lambda tth, wave: 2.0 * np.pi * npT2stl(tth, wave)  #=2pi*d*
ateln2 = 8.0 * math.log(2.0)
sateln2 = np.sqrt(ateln2)
nxs = np.newaxis

################################################################################
#### Powder utilities
################################################################################


[docs]def PhaseWtSum(G2frame, histo): ''' Calculate sum of phase mass*phase fraction for PWDR data (exclude magnetic phases) :param G2frame: GSASII main frame structure :param str histo: histogram name :returns: sum(scale*mass) for phases in histo ''' Histograms, Phases = G2frame.GetUsedHistogramsAndPhasesfromTree() wtSum = 0.0 for phase in Phases: if Phases[phase]['General']['Type'] != 'magnetic': if histo in Phases[phase]['Histograms']: if not Phases[phase]['Histograms'][histo]['Use']: continue mass = Phases[phase]['General']['Mass'] phFr = Phases[phase]['Histograms'][histo]['Scale'][0] wtSum += mass * phFr return wtSum
################################################################################ #GSASII pdf calculation routines ################################################################################
[docs]def Transmission(Geometry, Abs, Diam): ''' Calculate sample transmission :param str Geometry: one of 'Cylinder','Bragg-Brentano','Tilting flat plate in transmission','Fixed flat plate' :param float Abs: absorption coeff in cm-1 :param float Diam: sample thickness/diameter in mm ''' if 'Cylinder' in Geometry: #Lobanov & Alte da Veiga for 2-theta = 0; beam fully illuminates sample MuR = Abs * Diam / 20.0 if MuR <= 3.0: T0 = 16 / (3. * math.pi) T1 = -0.045780 T2 = -0.02489 T3 = 0.003045 T = -T0 * MuR - T1 * MuR**2 - T2 * MuR**3 - T3 * MuR**4 if T < -20.: return 2.06e-9 else: return math.exp(T) else: T1 = 1.433902 T2 = 0.013869 + 0.337894 T3 = 1.933433 + 1.163198 T4 = 0.044365 - 0.04259 T = (T1 - T4) / (1.0 + T2 * (MuR - 3.0))**T3 + T4 return T / 100. elif 'plate' in Geometry: MuR = Abs * Diam / 10. return math.exp(-MuR) elif 'Bragg' in Geometry: return 0.0
[docs]def SurfaceRough(SRA, SRB, Tth): ''' Suortti (J. Appl. Cryst, 5,325-331, 1972) surface roughness correction :param float SRA: Suortti surface roughness parameter :param float SRB: Suortti surface roughness parameter :param float Tth: 2-theta(deg) - can be numpy array ''' sth = npsind(Tth / 2.) T1 = np.exp(-SRB / sth) T2 = SRA + (1. - SRA) * np.exp(-SRB) return (SRA + (1. - SRA) * T1) / T2
[docs]def SurfaceRoughDerv(SRA, SRB, Tth): ''' Suortti surface roughness correction derivatives :param float SRA: Suortti surface roughness parameter (dimensionless) :param float SRB: Suortti surface roughness parameter (dimensionless) :param float Tth: 2-theta(deg) - can be numpy array :return list: [dydSRA,dydSRB] derivatives to be used for intensity derivative ''' sth = npsind(Tth / 2.) T1 = np.exp(-SRB / sth) T2 = SRA + (1. - SRA) * np.exp(-SRB) Trans = (SRA + (1. - SRA) * T1) / T2 dydSRA = ((1. - T1) * T2 - (1. - np.exp(-SRB)) * Trans) / T2**2 dydSRB = ((SRA - 1.) * T1 * T2 / sth - Trans * (SRA - T2)) / T2**2 return [dydSRA, dydSRB]
[docs]def Absorb(Geometry, MuR, Tth, Phi=0, Psi=0): '''Calculate sample absorption :param str Geometry: one of 'Cylinder','Bragg-Brentano','Tilting Flat Plate in transmission','Fixed flat plate' :param float MuR: absorption coeff * sample thickness/2 or radius :param Tth: 2-theta scattering angle - can be numpy array :param float Phi: flat plate tilt angle - future :param float Psi: flat plate tilt axis - future ''' def muRunder3(MuR, Sth2): T0 = 16.0 / (3. * np.pi) T1 = (25.99978-0.01911*Sth2**0.25)*np.exp(-0.024551*Sth2)+ \ 0.109561*np.sqrt(Sth2)-26.04556 T2 = -0.02489-0.39499*Sth2+1.219077*Sth2**1.5- \ 1.31268*Sth2**2+0.871081*Sth2**2.5-0.2327*Sth2**3 T3 = 0.003045 + 0.018167 * Sth2 - 0.03305 * Sth2**2 Trns = -T0 * MuR - T1 * MuR**2 - T2 * MuR**3 - T3 * MuR**4 return np.exp(Trns) def muRover3(MuR, Sth2): T1 = 1.433902+11.07504*Sth2-8.77629*Sth2*Sth2+ \ 10.02088*Sth2**3-3.36778*Sth2**4 T2 = (0.013869-0.01249*Sth2)*np.exp(3.27094*Sth2)+ \ (0.337894+13.77317*Sth2)/(1.0+11.53544*Sth2)**1.555039 T3 = 1.933433/(1.0+23.12967*Sth2)**1.686715- \ 0.13576*np.sqrt(Sth2)+1.163198 T4 = 0.044365 - 0.04259 / (1.0 + 0.41051 * Sth2)**148.4202 Trns = (T1 - T4) / (1.0 + T2 * (MuR - 3.0))**T3 + T4 return Trns / 100. Sth2 = npsind(Tth / 2.0)**2 if 'Cylinder' in Geometry: #Lobanov & Alte da Veiga for 2-theta = 0; beam fully illuminates sample if 'array' in str(type(MuR)): MuRSTh2 = np.concatenate((MuR, Sth2)) AbsCr = np.where(MuRSTh2[0] <= 3.0, muRunder3(MuRSTh2[0], MuRSTh2[1]), muRover3(MuRSTh2[0], MuRSTh2[1])) return AbsCr else: if MuR <= 3.0: return muRunder3(MuR, Sth2) else: return muRover3(MuR, Sth2) elif 'Bragg' in Geometry: return 1.0 elif 'Fixed' in Geometry: #assumes sample plane is perpendicular to incident beam # and only defined for 2theta < 90 MuT = 2. * MuR T1 = np.exp(-MuT) T2 = np.exp(-MuT / npcosd(Tth)) Tb = MuT - MuT / npcosd(Tth) return (T2 - T1) / Tb elif 'Tilting' in Geometry: #assumes symmetric tilt so sample plane is parallel to diffraction vector MuT = 2. * MuR cth = npcosd(Tth / 2.0) return np.exp(-MuT / cth) / cth
[docs]def AbsorbDerv(Geometry, MuR, Tth, Phi=0, Psi=0): 'needs a doc string' dA = 0.001 AbsP = Absorb(Geometry, MuR + dA, Tth, Phi, Psi) if MuR: AbsM = Absorb(Geometry, MuR - dA, Tth, Phi, Psi) return (AbsP - AbsM) / (2.0 * dA) else: return (AbsP - 1.) / dA
[docs]def Polarization(Pola, Tth, Azm=0.0): """ Calculate angle dependent x-ray polarization correction (not scaled correctly!) :param Pola: polarization coefficient e.g 1.0 fully polarized, 0.5 unpolarized :param Azm: azimuthal angle e.g. 0.0 in plane of polarization :param Tth: 2-theta scattering angle - can be numpy array which (if either) of these is "right"? :return: (pola, dpdPola) * pola = ((1-Pola)*npcosd(Azm)**2+Pola*npsind(Azm)**2)*npcosd(Tth)**2+ \ (1-Pola)*npsind(Azm)**2+Pola*npcosd(Azm)**2 * dpdPola: derivative needed for least squares """ pola = ((1.0-Pola)*npcosd(Azm)**2+Pola*npsind(Azm)**2)*npcosd(Tth)**2+ \ (1.0-Pola)*npsind(Azm)**2+Pola*npcosd(Azm)**2 dpdPola = -npsind(Tth)**2 * (npsind(Azm)**2 - npcosd(Azm)**2) return pola, dpdPola
[docs]def Oblique(ObCoeff, Tth): 'currently assumes detector is normal to beam' if ObCoeff: return (1. - ObCoeff) / (1.0 - np.exp(np.log(ObCoeff) / npcosd(Tth))) else: return 1.0
[docs]def Ruland(RulCoff, wave, Q, Compton): 'needs a doc string' C = 2.9978e8 D = 1.5e-3 hmc = 0.024262734687 sinth2 = (Q * wave / (4.0 * np.pi))**2 dlam = (wave**2) * Compton * Q / C dlam_c = 2.0 * hmc * sinth2 - D * wave**2 return 1.0 / ((1.0 + dlam / RulCoff) * (1.0 + (np.pi * dlam_c / (dlam + RulCoff))**2))
[docs]def LorchWeight(Q): 'needs a doc string' return np.sin(np.pi * (Q[-1] - Q) / (2.0 * Q[-1]))
[docs]def GetAsfMean(ElList, Sthl2): '''Calculate various scattering factor terms for PDF calcs :param dict ElList: element dictionary contains scattering factor coefficients, etc. :param np.array Sthl2: numpy array of sin theta/lambda squared values :returns: mean(f^2), mean(f)^2, mean(compton) ''' sumNoAtoms = 0.0 FF = np.zeros_like(Sthl2) FF2 = np.zeros_like(Sthl2) CF = np.zeros_like(Sthl2) for El in ElList: sumNoAtoms += ElList[El]['FormulaNo'] for El in ElList: el = ElList[El] ff2 = (G2elem.ScatFac(el, Sthl2) + el['fp'])**2 + el['fpp']**2 cf = G2elem.ComptonFac(el, Sthl2) FF += np.sqrt(ff2) * el['FormulaNo'] / sumNoAtoms FF2 += ff2 * el['FormulaNo'] / sumNoAtoms CF += cf * el['FormulaNo'] / sumNoAtoms return FF2, FF**2, CF
[docs]def GetNumDensity(ElList, Vol): 'needs a doc string' sumNoAtoms = 0.0 for El in ElList: sumNoAtoms += ElList[El]['FormulaNo'] return sumNoAtoms / Vol
[docs]def CalcPDF(data, inst, limits, xydata): '''Computes I(Q), S(Q) & G(r) from Sample, Bkg, etc. diffraction patterns loaded into dict xydata; results are placed in xydata. Calculation parameters are found in dicts data and inst and list limits. The return value is at present an empty list. ''' auxPlot = [] Ibeg = np.searchsorted(xydata['Sample'][1][0], limits[0]) Ifin = np.searchsorted(xydata['Sample'][1][0], limits[1]) + 1 #subtract backgrounds - if any & use PWDR limits # GSASIIpath.IPyBreak() IofQ = copy.deepcopy(xydata['Sample']) IofQ[1] = np.array(IofQ[1])[:, Ibeg:Ifin] if data['Sample Bkg.']['Name']: IofQ[1][1] += xydata['Sample Bkg.'][1][1][Ibeg:Ifin] * data[ 'Sample Bkg.']['Mult'] if data['Container']['Name']: xycontainer = xydata['Container'][1][1] * data['Container']['Mult'] if data['Container Bkg.']['Name']: xycontainer += xydata['Container Bkg.'][1][1][Ibeg:Ifin] * data[ 'Container Bkg.']['Mult'] IofQ[1][1] += xycontainer[Ibeg:Ifin] data['IofQmin'] = IofQ[1][1][-1] IofQ[1][1] -= data.get('Flat Bkg', 0.) #get element data & absorption coeff. ElList = data['ElList'] Abs = G2lat.CellAbsorption(ElList, data['Form Vol']) #Apply angle dependent corrections Tth = IofQ[1][0] MuR = Abs * data['Diam'] / 20.0 IofQ[1][1] /= Absorb(data['Geometry'], MuR, Tth) if 'X' in inst['Type'][0]: IofQ[1][1] /= Polarization(inst['Polariz.'][1], Tth, Azm=inst['Azimuth'][1])[0] if data['DetType'] == 'Image plate': IofQ[1][1] *= Oblique(data['ObliqCoeff'], Tth) XY = IofQ[1] #convert to Q # nQpoints = len(XY[0]) #points for Q interpolation nQpoints = 5000 if 'C' in inst['Type'][0]: wave = G2mth.getWave(inst) minQ = npT2q(Tth[0], wave) maxQ = npT2q(Tth[-1], wave) Qpoints = np.linspace(0., maxQ, nQpoints, endpoint=True) dq = Qpoints[1] - Qpoints[0] XY[0] = npT2q(XY[0], wave) elif 'T' in inst['Type'][0]: difC = inst['difC'][1] minQ = 2. * np.pi * difC / Tth[-1] maxQ = 2. * np.pi * difC / Tth[0] Qpoints = np.linspace(0., maxQ, nQpoints, endpoint=True) dq = Qpoints[1] - Qpoints[0] XY[0] = 2. * np.pi * difC / XY[0] Qdata = si.griddata(XY[0], XY[1], Qpoints, method='linear', fill_value=XY[1][0]) #interpolate I(Q) Qdata -= np.min(Qdata) * data['BackRatio'] qLimits = data['QScaleLim'] minQ = np.searchsorted(Qpoints, qLimits[0]) maxQ = np.searchsorted(Qpoints, qLimits[1]) + 1 newdata = [] if len(IofQ) < 3: xydata['IofQ'] = [IofQ[0], [Qpoints, Qdata], ''] else: xydata['IofQ'] = [IofQ[0], [Qpoints, Qdata], IofQ[2]] for item in xydata['IofQ'][1]: newdata.append(item[:maxQ]) xydata['IofQ'][1] = newdata xydata['SofQ'] = copy.deepcopy(xydata['IofQ']) FFSq, SqFF, CF = GetAsfMean(ElList, (xydata['SofQ'][1][0] / (4.0 * np.pi))**2) #these are <f^2>,<f>^2,Cf Q = xydata['SofQ'][1][0] # auxPlot.append([Q,np.copy(CF),'CF-unCorr']) ruland = Ruland(data['Ruland'], wave, Q, CF) # auxPlot.append([Q,ruland,'Ruland']) CF *= ruland # auxPlot.append([Q,CF,'CF-Corr']) scale = np.sum( (FFSq + CF)[minQ:maxQ]) / np.sum(xydata['SofQ'][1][1][minQ:maxQ]) xydata['SofQ'][1][1] *= scale xydata['SofQ'][1][1] -= CF xydata['SofQ'][1][1] = xydata['SofQ'][1][1] / SqFF scale = len(xydata['SofQ'][1][1][minQ:maxQ]) / np.sum( xydata['SofQ'][1][1][minQ:maxQ]) xydata['SofQ'][1][1] *= scale xydata['FofQ'] = copy.deepcopy(xydata['SofQ']) xydata['FofQ'][1][1] = xydata['FofQ'][1][0] * (xydata['SofQ'][1][1] - 1.0) if data['Lorch']: xydata['FofQ'][1][1] *= LorchWeight(Q) xydata['GofR'] = copy.deepcopy(xydata['FofQ']) nR = len(xydata['GofR'][1][1]) Rmax = GSASIIpath.GetConfigValue('PDF_Rmax', 100.) mul = int(round(2. * np.pi * nR / (Rmax * qLimits[1]))) # mul = int(round(2.*np.pi*nR/(data.get('Rmax',100.)*qLimits[1]))) xydata['GofR'][1][0] = 2. * np.pi * np.linspace( 0, nR, nR, endpoint=True) / (mul * qLimits[1]) xydata['GofR'][1][1] = -dq * np.imag( fft.fft(xydata['FofQ'][1][1], mul * nR)[:nR]) if data.get('noRing', True): xydata['GofR'][1][1] = np.where(xydata['GofR'][1][0] < 0.5, 0., xydata['GofR'][1][1]) return auxPlot
[docs]def PDFPeakFit(peaks, data): rs2pi = 1. / np.sqrt(2 * np.pi) def MakeParms(peaks): varyList = [] parmDict = {'slope': peaks['Background'][1][1]} if peaks['Background'][2]: varyList.append('slope') for i, peak in enumerate(peaks['Peaks']): parmDict['PDFpos;' + str(i)] = peak[0] parmDict['PDFmag;' + str(i)] = peak[1] parmDict['PDFsig;' + str(i)] = peak[2] if 'P' in peak[3]: varyList.append('PDFpos;' + str(i)) if 'M' in peak[3]: varyList.append('PDFmag;' + str(i)) if 'S' in peak[3]: varyList.append('PDFsig;' + str(i)) return parmDict, varyList def SetParms(peaks, parmDict, varyList): if 'slope' in varyList: peaks['Background'][1][1] = parmDict['slope'] for i, peak in enumerate(peaks['Peaks']): if 'PDFpos;' + str(i) in varyList: peak[0] = parmDict['PDFpos;' + str(i)] if 'PDFmag;' + str(i) in varyList: peak[1] = parmDict['PDFmag;' + str(i)] if 'PDFsig;' + str(i) in varyList: peak[2] = parmDict['PDFsig;' + str(i)] def CalcPDFpeaks(parmdict, Xdata): Z = parmDict['slope'] * Xdata ipeak = 0 while True: try: pos = parmdict['PDFpos;' + str(ipeak)] mag = parmdict['PDFmag;' + str(ipeak)] wid = parmdict['PDFsig;' + str(ipeak)] wid2 = 2. * wid**2 Z += mag * rs2pi * np.exp(-(Xdata - pos)**2 / wid2) / wid ipeak += 1 except KeyError: #no more peaks to process return Z def errPDFProfile(values, xdata, ydata, parmdict, varylist): parmdict.update(zip(varylist, values)) M = CalcPDFpeaks(parmdict, xdata) - ydata return M newpeaks = copy.copy(peaks) iBeg = np.searchsorted(data[1][0], newpeaks['Limits'][0]) iFin = np.searchsorted(data[1][0], newpeaks['Limits'][1]) + 1 X = data[1][0][iBeg:iFin] Y = data[1][1][iBeg:iFin] parmDict, varyList = MakeParms(peaks) if not len(varyList): G2fil.G2Print(' Nothing varied') return newpeaks, None, None, None, None, None Rvals = {} values = np.array(Dict2Values(parmDict, varyList)) result = so.leastsq(errPDFProfile, values, full_output=True, ftol=0.0001, args=(X, Y, parmDict, varyList)) chisq = np.sum(result[2]['fvec']**2) Values2Dict(parmDict, varyList, result[0]) SetParms(peaks, parmDict, varyList) Rvals['Rwp'] = np.sqrt(chisq / np.sum(Y**2)) * 100. #to % chisq = np.sum(result[2]['fvec']**2) / (len(X) - len(values) ) #reduced chi^2 = M/(Nobs-Nvar) sigList = list(np.sqrt(chisq * np.diag(result[1]))) Z = CalcPDFpeaks(parmDict, X) newpeaks['calc'] = [X, Z] return newpeaks, result[0], varyList, sigList, parmDict, Rvals
[docs]def MakeRDF(RDFcontrols, background, inst, pwddata): import scipy.signal as signal auxPlot = [] if 'C' in inst['Type'][0]: Tth = pwddata[0] wave = G2mth.getWave(inst) minQ = npT2q(Tth[0], wave) maxQ = npT2q(Tth[-1], wave) powQ = npT2q(Tth, wave) elif 'T' in inst['Type'][0]: TOF = pwddata[0] difC = inst['difC'][1] minQ = 2. * np.pi * difC / TOF[-1] maxQ = 2. * np.pi * difC / TOF[0] powQ = 2. * np.pi * difC / TOF piDQ = np.pi / (maxQ - minQ) Qpoints = np.linspace(minQ, maxQ, len(pwddata[0]), endpoint=True) if RDFcontrols['UseObsCalc'] == 'obs-calc': Qdata = si.griddata(powQ, pwddata[1] - pwddata[3], Qpoints, method=RDFcontrols['Smooth'], fill_value=0.) elif RDFcontrols['UseObsCalc'] == 'obs-back': Qdata = si.griddata(powQ, pwddata[1] - pwddata[4], Qpoints, method=RDFcontrols['Smooth'], fill_value=pwddata[1][0]) elif RDFcontrols['UseObsCalc'] == 'calc-back': Qdata = si.griddata(powQ, pwddata[3] - pwddata[4], Qpoints, method=RDFcontrols['Smooth'], fill_value=pwddata[1][0]) Qdata *= np.sin((Qpoints - minQ) * piDQ) / piDQ Qdata *= 0.5 * np.sqrt(Qpoints) #Qbin normalization # GSASIIpath.IPyBreak() dq = Qpoints[1] - Qpoints[0] nR = len(Qdata) R = 0.5 * np.pi * np.linspace(0, nR, nR) / (4. * maxQ) iFin = np.searchsorted(R, RDFcontrols['maxR']) + 1 bBut, aBut = signal.butter(4, 0.01) Qsmooth = signal.filtfilt(bBut, aBut, Qdata) # auxPlot.append([Qpoints,Qdata,'interpolate:'+RDFcontrols['Smooth']]) # auxPlot.append([Qpoints,Qsmooth,'interpolate:'+RDFcontrols['Smooth']]) DofR = dq * np.imag(fft.fft(Qsmooth, 16 * nR)[:nR]) # DofR = dq*np.imag(ft.fft(Qsmooth,16*nR)[:nR]) auxPlot.append( [R[:iFin], DofR[:iFin], 'D(R) for ' + RDFcontrols['UseObsCalc']]) return auxPlot
# PDF optimization =============================================================
[docs]def OptimizePDF(data, xydata, limits, inst, showFit=True, maxCycles=5): import scipy.optimize as opt numbDen = GetNumDensity(data['ElList'], data['Form Vol']) Min, Init, Done = SetupPDFEval(data, xydata, limits, inst, numbDen) xstart = Init() bakMul = data['Sample Bkg.']['Mult'] if showFit: rms = Min(xstart) G2fil.G2Print(' Optimizing corrections to improve G(r) at low r') if data['Sample Bkg.'].get('Refine', False): # data['Flat Bkg'] = 0. G2fil.G2Print( ' start: Ruland={:.3f}, Sample Bkg mult={:.3f} (RMS:{:.4f})'. format(data['Ruland'], data['Sample Bkg.']['Mult'], rms)) else: G2fil.G2Print( ' start: Flat Bkg={:.1f}, BackRatio={:.3f}, Ruland={:.3f} (RMS:{:.4f})' .format(data['Flat Bkg'], data['BackRatio'], data['Ruland'], rms)) if data['Sample Bkg.'].get('Refine', False): res = opt.minimize(Min, xstart, bounds=([0.01, 1], [1.2 * bakMul, 0.8 * bakMul]), method='L-BFGS-B', options={'maxiter': maxCycles}, tol=0.001) else: res = opt.minimize(Min, xstart, bounds=([0, None], [0, 1], [0.01, 1]), method='L-BFGS-B', options={'maxiter': maxCycles}, tol=0.001) Done(res['x']) if showFit: if res['success']: msg = 'Converged' else: msg = 'Not Converged' if data['Sample Bkg.'].get('Refine', False): G2fil.G2Print( ' end: Ruland={:.3f}, Sample Bkg mult={:.3f} (RMS:{:.4f}) *** {} ***\n' .format(data['Ruland'], data['Sample Bkg.']['Mult'], res['fun'], msg)) else: G2fil.G2Print( ' end: Flat Bkg={:.1f}, BackRatio={:.3f}, Ruland={:.3f}) *** {} ***\n' .format(data['Flat Bkg'], data['BackRatio'], data['Ruland'], res['fun'], msg)) return res
[docs]def SetupPDFEval(data, xydata, limits, inst, numbDen): Data = copy.deepcopy(data) BkgMax = 1. def EvalLowPDF(arg): '''Objective routine -- evaluates the RMS deviations in G(r) from -4(pi)*#density*r for for r<Rmin arguments are ['Flat Bkg','BackRatio','Ruland'] scaled so that the min & max values are between 0 and 1. ''' if Data['Sample Bkg.'].get('Refine', False): R, S = arg Data['Sample Bkg.']['Mult'] = S else: F, B, R = arg Data['Flat Bkg'] = F * BkgMax Data['BackRatio'] = B Data['Ruland'] = R / 10. CalcPDF(Data, inst, limits, xydata) # test low r computation g = xydata['GofR'][1][1] r = xydata['GofR'][1][0] g0 = g[r < Data['Rmin']] + 4 * np.pi * r[r < Data['Rmin']] * numbDen M = sum(g0**2) / len(g0) return M def GetCurrentVals(): '''Get the current ['Flat Bkg','BackRatio','Ruland'] with scaling ''' if data['Sample Bkg.'].get('Refine', False): return [max(10 * data['Ruland'], .05), data['Sample']['Mult']] try: F = data['Flat Bkg'] / BkgMax except: F = 0 return [F, data['BackRatio'], max(10 * data['Ruland'], .05)] def SetFinalVals(arg): '''Set the 'Flat Bkg', 'BackRatio' & 'Ruland' values from the scaled, refined values and plot corrected region of G(r) ''' if data['Sample Bkg.'].get('Refine', False): R, S = arg data['Sample Bkg.']['Mult'] = S else: F, B, R = arg data['Flat Bkg'] = F * BkgMax data['BackRatio'] = B data['Ruland'] = R / 10. CalcPDF(data, inst, limits, xydata) EvalLowPDF(GetCurrentVals()) BkgMax = max(xydata['IofQ'][1][1]) / 50. return EvalLowPDF, GetCurrentVals, SetFinalVals
################################################################################ #GSASII peak fitting routines: Finger, Cox & Jephcoat model ################################################################################
[docs]def factorize(num): ''' Provide prime number factors for integer num :returns: dictionary of prime factors (keys) & power for each (data) ''' factors = {} orig = num # we take advantage of the fact that (i +1)**2 = i**2 + 2*i +1 i, sqi = 2, 4 while sqi <= num: while not num % i: num /= i factors[i] = factors.get(i, 0) + 1 sqi += 2 * i + 1 i += 1 if num != 1 and num != orig: factors[num] = factors.get(num, 0) + 1 if factors: return factors else: return {num: 1} #a prime number!
[docs]def makeFFTsizeList(nmin=1, nmax=1023, thresh=15): ''' Provide list of optimal data sizes for FFT calculations :param int nmin: minimum data size >= 1 :param int nmax: maximum data size > nmin :param int thresh: maximum prime factor allowed :Returns: list of data sizes where the maximum prime factor is < thresh ''' plist = [] nmin = max(1, nmin) nmax = max(nmin + 1, nmax) for p in range(nmin, nmax): if max(list(factorize(p).keys())) < thresh: plist.append(p) return plist
np.seterr(divide='ignore') # Normal distribution # loc = mu, scale = std _norm_pdf_C = 1. / math.sqrt(2 * math.pi)
[docs]class norm_gen(st.rv_continuous): 'needs a doc string'
[docs] def pdf(self, x, *args, **kwds): loc, scale = kwds['loc'], kwds['scale'] x = (x - loc) / scale return np.exp(-x**2 / 2.0) * _norm_pdf_C / scale
norm = norm_gen(name='norm', longname='A normal', extradoc=""" Normal distribution The location (loc) keyword specifies the mean. The scale (scale) keyword specifies the standard deviation. normal.pdf(x) = exp(-x**2/2)/sqrt(2*pi) """) ## Cauchy # median = loc
[docs]class cauchy_gen(st.rv_continuous): 'needs a doc string'
[docs] def pdf(self, x, *args, **kwds): loc, scale = kwds['loc'], kwds['scale'] x = (x - loc) / scale return 1.0 / np.pi / (1.0 + x * x) / scale
cauchy = cauchy_gen(name='cauchy', longname='Cauchy', extradoc=""" Cauchy distribution cauchy.pdf(x) = 1/(pi*(1+x**2)) This is the t distribution with one degree of freedom. """) #GSASII peak fitting routine: Finger, Cox & Jephcoat model
[docs]class fcjde_gen(st.rv_continuous): """ Finger-Cox-Jephcoat D(2phi,2th) function for S/L = H/L Ref: J. Appl. Cryst. (1994) 27, 892-900. :param x: array -1 to 1 :param t: 2-theta position of peak :param s: sum(S/L,H/L); S: sample height, H: detector opening, L: sample to detector opening distance :param dx: 2-theta step size in deg :returns: for fcj.pdf * T = x*dx+t * s = S/L+H/L * if x < 0:: fcj.pdf = [1/sqrt({cos(T)**2/cos(t)**2}-1) - 1/s]/|cos(T)| * if x >= 0: fcj.pdf = 0 """ def _pdf(self, x, t, s, dx): T = dx * x + t ax2 = abs(npcosd(T)) ax = ax2**2 bx = npcosd(t)**2 bx = np.where(ax > bx, bx, ax) fx = np.where(ax > bx, (np.sqrt(bx / (ax - bx)) - 1. / s) / ax2, 0.0) fx = np.where(fx > 0., fx, 0.0) return fx
[docs] def pdf(self, x, *args, **kwds): loc = kwds['loc'] return self._pdf(x - loc, *args)
fcjde = fcjde_gen(name='fcjde', shapes='t,s,dx')
[docs]def getWidthsCW(pos, sig, gam, shl): '''Compute the peak widths used for computing the range of a peak for constant wavelength data. On low-angle side, 50 FWHM are used, on high-angle side 75 are used, low angle side extended for axial divergence (for peaks above 90 deg, these are reversed.) ''' widths = [np.sqrt(sig) / 100., gam / 100.] fwhm = 2.355 * widths[0] + widths[1] fmin = 50. * (fwhm + shl * abs(npcosd(pos))) fmax = 75.0 * fwhm if pos > 90: fmin, fmax = [fmax, fmin] return widths, fmin, fmax
[docs]def getWidthsTOF(pos, alp, bet, sig, gam): '''Compute the peak widths used for computing the range of a peak for constant wavelength data. 50 FWHM are used on both sides each extended by exponential coeff. ''' widths = [np.sqrt(sig), gam] fwhm = 2.355 * widths[0] + 2. * widths[1] fmin = 50. * fwhm * (1. + 1. / alp) fmax = 50. * fwhm * (1. + 1. / bet) return widths, fmin, fmax
[docs]def getFWHM(pos, Inst): '''Compute total FWHM from Thompson, Cox & Hastings (1987) , J. Appl. Cryst. 20, 79-83 via getgamFW(g,s). :param pos: float peak position in deg 2-theta or tof in musec :param Inst: dict instrument parameters :returns float: total FWHM of pseudoVoigt in deg or musec ''' sig = lambda Th, U, V, W: np.sqrt( max(0.001, U * tand(Th)**2 + V * tand(Th) + W)) sigTOF = lambda dsp, S0, S1, S2, Sq: np.sqrt(S0 + S1 * dsp**2 + S2 * dsp**4 + Sq * dsp) gam = lambda Th, X, Y, Z: Z + X / cosd(Th) + Y * tand(Th) gamTOF = lambda dsp, X, Y, Z: Z + X * dsp + Y * dsp**2 alpTOF = lambda dsp, alp: alp / dsp betTOF = lambda dsp, bet0, bet1, betq: bet0 + bet1 / dsp**4 + betq / dsp**2 if 'C' in Inst['Type'][0]: s = sig(pos / 2., Inst['U'][1], Inst['V'][1], Inst['W'][1]) g = gam(pos / 2., Inst['X'][1], Inst['Y'][1], Inst['Z'][1]) return getgamFW(g, s) / 100. #returns FWHM in deg else: dsp = pos / Inst['difC'][0] alp = alpTOF(dsp, Inst['alpha'][0]) bet = betTOF(dsp, Inst['beta-0'][0], Inst['beta-1'][0], Inst['beta-q'][0]) s = sigTOF(dsp, Inst['sig-0'][1], Inst['sig-1'][1], Inst['sig-2'][1], Inst['sig-q'][1]) g = gamTOF(dsp, Inst['X'][1], Inst['Y'][1], Inst['Z'][1]) return getgamFW(g, s) + np.log(2.0) * (alp + bet) / (alp * bet)
[docs]def getgamFW(g, s): '''Compute total FWHM from Thompson, Cox & Hastings (1987), J. Appl. Cryst. 20, 79-83 lambda fxn needs FWHM for both Gaussian & Lorentzian components :param g: float Lorentzian gamma = FWHM(L) :param s: float Gaussian sig :returns float: total FWHM of pseudoVoigt ''' gamFW = lambda s, g: np.exp( np.log(s**5 + 2.69269 * s**4 * g + 2.42843 * s**3 * g**2 + 4.47163 * s** 2 * g**3 + 0.07842 * s * g**4 + g**5) / 5.) return gamFW(2.35482 * s, g) #sqrt(8ln2)*sig = FWHM(G)
[docs]def getFCJVoigt(pos, intens, sig, gam, shl, xdata): '''Compute the Finger-Cox-Jepcoat modified Voigt function for a CW powder peak by direct convolution. This version is not used. ''' DX = xdata[1] - xdata[0] widths, fmin, fmax = getWidthsCW(pos, sig, gam, shl) x = np.linspace(pos - fmin, pos + fmin, 256) dx = x[1] - x[0] Norm = norm.pdf(x, loc=pos, scale=widths[0]) Cauchy = cauchy.pdf(x, loc=pos, scale=widths[1]) arg = [ pos, shl / 57.2958, dx, ] FCJ = fcjde.pdf(x, *arg, loc=pos) if len(np.nonzero(FCJ)[0]) > 5: z = np.column_stack([Norm, Cauchy, FCJ]).T Z = fft.fft(z) Df = fft.ifft(Z.prod(axis=0)).real else: z = np.column_stack([Norm, Cauchy]).T Z = fft.fft(z) Df = fft.fftshift(fft.ifft(Z.prod(axis=0))).real Df /= np.sum(Df) Df = si.interp1d(x, Df, bounds_error=False, fill_value=0.0) return intens * Df(xdata) * DX / dx
[docs]def getBackground(pfx, parmDict, bakType, dataType, xdata, fixedBkg={}): '''Computes the background from vars pulled from gpx file or tree. ''' if 'T' in dataType: q = 2. * np.pi * parmDict[pfx + 'difC'] / xdata elif 'C' in dataType: wave = parmDict.get(pfx + 'Lam', parmDict.get(pfx + 'Lam1', 1.0)) q = npT2q(xdata, wave) yb = np.zeros_like(xdata) nBak = 0 cw = np.diff(xdata) cw = np.append(cw, cw[-1]) sumBk = [0., 0., 0] while True: key = pfx + 'Back;' + str(nBak) if key in parmDict: nBak += 1 else: break #empirical functions if bakType in ['chebyschev', 'cosine']: dt = xdata[-1] - xdata[0] for iBak in range(nBak): key = pfx + 'Back;' + str(iBak) if bakType == 'chebyschev': ybi = parmDict[key] * (-1. + 2. * (xdata - xdata[0]) / dt)**iBak elif bakType == 'cosine': ybi = parmDict[key] * npcosd(180. * xdata * iBak / xdata[-1]) yb += ybi sumBk[0] = np.sum(yb) elif bakType in ['Q^2 power series', 'Q^-2 power series']: QT = 1. yb += np.ones_like(yb) * parmDict[pfx + 'Back;0'] for iBak in range(nBak - 1): key = pfx + 'Back;' + str(iBak + 1) if '-2' in bakType: QT *= (iBak + 1) * q**-2 else: QT *= q**2 / (iBak + 1) yb += QT * parmDict[key] sumBk[0] = np.sum(yb) elif bakType in [ 'lin interpolate', 'inv interpolate', 'log interpolate', ]: if nBak == 1: yb = np.ones_like(xdata) * parmDict[pfx + 'Back;0'] elif nBak == 2: dX = xdata[-1] - xdata[0] T2 = (xdata - xdata[0]) / dX T1 = 1.0 - T2 yb = parmDict[pfx + 'Back;0'] * T1 + parmDict[pfx + 'Back;1'] * T2 else: xnomask = ma.getdata(xdata) xmin, xmax = xnomask[0], xnomask[-1] if bakType == 'lin interpolate': bakPos = np.linspace(xmin, xmax, nBak, True) elif bakType == 'inv interpolate': bakPos = 1. / np.linspace(1. / xmax, 1. / xmin, nBak, True) elif bakType == 'log interpolate': bakPos = np.exp( np.linspace(np.log(xmin), np.log(xmax), nBak, True)) bakPos[0] = xmin bakPos[-1] = xmax bakVals = np.zeros(nBak) for i in range(nBak): bakVals[i] = parmDict[pfx + 'Back;' + str(i)] bakInt = si.interp1d(bakPos, bakVals, 'linear') yb = bakInt(ma.getdata(xdata)) sumBk[0] = np.sum(yb) #Debye function if pfx + 'difC' in parmDict: ff = 1. else: try: wave = parmDict[pfx + 'Lam'] except KeyError: wave = parmDict[pfx + 'Lam1'] SQ = (q / (4. * np.pi))**2 FF = G2elem.GetFormFactorCoeff('Si')[0] ff = np.array(G2elem.ScatFac(FF, SQ)[0])**2 iD = 0 while True: try: dbA = parmDict[pfx + 'DebyeA;' + str(iD)] dbR = parmDict[pfx + 'DebyeR;' + str(iD)] dbU = parmDict[pfx + 'DebyeU;' + str(iD)] ybi = ff * dbA * np.sin(q * dbR) * np.exp(-dbU * q**2) / (q * dbR) yb += ybi sumBk[1] += np.sum(ybi) iD += 1 except KeyError: break #peaks iD = 0 while True: try: pkP = parmDict[pfx + 'BkPkpos;' + str(iD)] pkI = parmDict[pfx + 'BkPkint;' + str(iD)] pkS = parmDict[pfx + 'BkPksig;' + str(iD)] pkG = parmDict[pfx + 'BkPkgam;' + str(iD)] if 'C' in dataType: Wd, fmin, fmax = getWidthsCW(pkP, pkS, pkG, .002) else: #'T'OF Wd, fmin, fmax = getWidthsTOF(pkP, 1., 1., pkS, pkG) iBeg = np.searchsorted(xdata, pkP - fmin) iFin = np.searchsorted(xdata, pkP + fmax) lenX = len(xdata) if not iBeg: iFin = np.searchsorted(xdata, pkP + fmax) elif iBeg == lenX: iFin = iBeg else: iFin = np.searchsorted(xdata, pkP + fmax) if 'C' in dataType: ybi = pkI * getFCJVoigt3(pkP, pkS, pkG, 0.002, xdata[iBeg:iFin]) yb[iBeg:iFin] += ybi else: #'T'OF ybi = pkI * getEpsVoigt(pkP, 1., 1., pkS, pkG, xdata[iBeg:iFin]) yb[iBeg:iFin] += ybi sumBk[2] += np.sum(ybi) iD += 1 except KeyError: break except ValueError: G2fil.G2Print('**** WARNING - backround peak ' + str(iD) + ' sigma is negative; fix & try again ****') break # fixed background from file if len(fixedBkg) >= 3: mult = fixedBkg.get('_fixedMult', 0.0) if len(fixedBkg.get('_fixedValues', [])) != len(yb): G2fil.G2Print( 'Lengths of backgrounds do not agree: yb={}, fixed={}'.format( len(yb), len(fixedBkg.get('_fixedValues', [])))) elif mult: yb -= mult * fixedBkg.get('_fixedValues', []) # N.B. mult is negative sumBk[0] = sum(yb) return yb, sumBk
[docs]def getBackgroundDerv(hfx, parmDict, bakType, dataType, xdata): 'needs a doc string' if 'T' in dataType: q = 2. * np.pi * parmDict[hfx + 'difC'] / xdata elif 'C' in dataType: wave = parmDict.get(hfx + 'Lam', parmDict.get(hfx + 'Lam1', 1.0)) q = 2. * np.pi * npsind(xdata / 2.) / wave nBak = 0 while True: key = hfx + 'Back;' + str(nBak) if key in parmDict: nBak += 1 else: break dydb = np.zeros(shape=(nBak, len(xdata))) dyddb = np.zeros(shape=(3 * parmDict[hfx + 'nDebye'], len(xdata))) dydpk = np.zeros(shape=(4 * parmDict[hfx + 'nPeaks'], len(xdata))) cw = np.diff(xdata) cw = np.append(cw, cw[-1]) if bakType in ['chebyschev', 'cosine']: dt = xdata[-1] - xdata[0] for iBak in range(nBak): if bakType == 'chebyschev': dydb[iBak] = (-1. + 2. * (xdata - xdata[0]) / dt)**iBak elif bakType == 'cosine': dydb[iBak] = npcosd(180. * xdata * iBak / xdata[-1]) elif bakType in ['Q^2 power series', 'Q^-2 power series']: QT = 1. dydb[0] = np.ones_like(xdata) for iBak in range(nBak - 1): if '-2' in bakType: QT *= (iBak + 1) * q**-2 else: QT *= q**2 / (iBak + 1) dydb[iBak + 1] = QT elif bakType in [ 'lin interpolate', 'inv interpolate', 'log interpolate', ]: if nBak == 1: dydb[0] = np.ones_like(xdata) elif nBak == 2: dX = xdata[-1] - xdata[0] T2 = (xdata - xdata[0]) / dX T1 = 1.0 - T2 dydb = [T1, T2] else: xnomask = ma.getdata(xdata) xmin, xmax = xnomask[0], xnomask[-1] if bakType == 'lin interpolate': bakPos = np.linspace(xmin, xmax, nBak, True) elif bakType == 'inv interpolate': bakPos = 1. / np.linspace(1. / xmax, 1. / xmin, nBak, True) elif bakType == 'log interpolate': bakPos = np.exp( np.linspace(np.log(xmin), np.log(xmax), nBak, True)) bakPos[0] = xmin bakPos[-1] = xmax for i, pos in enumerate(bakPos): if i == 0: dydb[0] = np.where(xdata < bakPos[1], (bakPos[1] - xdata) / (bakPos[1] - bakPos[0]), 0.) elif i == len(bakPos) - 1: dydb[i] = np.where(xdata > bakPos[-2], (bakPos[-1] - xdata) / (bakPos[-1] - bakPos[-2]), 0.) else: dydb[i] = np.where( xdata > bakPos[i], np.where(xdata < bakPos[i + 1], (bakPos[i + 1] - xdata) / (bakPos[i + 1] - bakPos[i]), 0.), np.where(xdata > bakPos[i - 1], (xdata - bakPos[i - 1]) / (bakPos[i] - bakPos[i - 1]), 0.)) if hfx + 'difC' in parmDict: ff = 1. else: wave = parmDict.get(hfx + 'Lam', parmDict.get(hfx + 'Lam1', 1.0)) q = npT2q(xdata, wave) SQ = (q / (4 * np.pi))**2 FF = G2elem.GetFormFactorCoeff('Si')[0] ff = np.array(G2elem.ScatFac( FF, SQ)[0]) * np.pi**2 #needs pi^2~10. for cw data (why?) iD = 0 while True: try: if hfx + 'difC' in parmDict: q = 2 * np.pi * parmDict[hfx + 'difC'] / xdata dbA = parmDict[hfx + 'DebyeA;' + str(iD)] dbR = parmDict[hfx + 'DebyeR;' + str(iD)] dbU = parmDict[hfx + 'DebyeU;' + str(iD)] sqr = np.sin(q * dbR) / (q * dbR) cqr = np.cos(q * dbR) temp = np.exp(-dbU * q**2) dyddb[3 * iD] = ff * sqr * temp dyddb[3 * iD + 1] = ff * dbA * temp * (cqr - sqr) / (dbR) dyddb[3 * iD + 2] = -ff * dbA * sqr * temp * q**2 iD += 1 except KeyError: break iD = 0 while True: try: pkP = parmDict[hfx + 'BkPkpos;' + str(iD)] pkI = parmDict[hfx + 'BkPkint;' + str(iD)] pkS = parmDict[hfx + 'BkPksig;' + str(iD)] pkG = parmDict[hfx + 'BkPkgam;' + str(iD)] if 'C' in dataType: Wd, fmin, fmax = getWidthsCW(pkP, pkS, pkG, .002) else: #'T'OF Wd, fmin, fmax = getWidthsTOF(pkP, 1., 1., pkS, pkG) iBeg = np.searchsorted(xdata, pkP - fmin) iFin = np.searchsorted(xdata, pkP + fmax) lenX = len(xdata) if not iBeg: iFin = np.searchsorted(xdata, pkP + fmax) elif iBeg == lenX: iFin = iBeg else: iFin = np.searchsorted(xdata, pkP + fmax) if 'C' in dataType: Df, dFdp, dFds, dFdg, x = getdFCJVoigt3(pkP, pkS, pkG, .002, xdata[iBeg:iFin]) dydpk[4 * iD][iBeg:iFin] += 100. * cw[iBeg:iFin] * pkI * dFdp dydpk[4 * iD + 1][iBeg:iFin] += 100. * cw[iBeg:iFin] * Df dydpk[4 * iD + 2][iBeg:iFin] += 100. * cw[iBeg:iFin] * pkI * dFds dydpk[4 * iD + 3][iBeg:iFin] += 100. * cw[iBeg:iFin] * pkI * dFdg else: #'T'OF Df, dFdp, x, x, dFds, dFdg = getdEpsVoigt( pkP, 1., 1., pkS, pkG, xdata[iBeg:iFin]) dydpk[4 * iD][iBeg:iFin] += pkI * dFdp dydpk[4 * iD + 1][iBeg:iFin] += Df dydpk[4 * iD + 2][iBeg:iFin] += pkI * dFds dydpk[4 * iD + 3][iBeg:iFin] += pkI * dFdg iD += 1 except KeyError: break except ValueError: G2fil.G2Print('**** WARNING - backround peak ' + str(iD) + ' sigma is negative; fix & try again ****') break return dydb, dyddb, dydpk
#use old fortran routine
[docs]def getFCJVoigt3(pos, sig, gam, shl, xdata): '''Compute the Finger-Cox-Jepcoat modified Pseudo-Voigt function for a CW powder peak in external Fortran routine ''' Df = pyd.pypsvfcj(len(xdata), xdata - pos, pos, sig, gam, shl) # Df = pyd.pypsvfcjo(len(xdata),xdata-pos,pos,sig,gam,shl) Df /= np.sum(Df) return Df
[docs]def getdFCJVoigt3(pos, sig, gam, shl, xdata): '''Compute analytic derivatives the Finger-Cox-Jepcoat modified Pseudo-Voigt function for a CW powder peak ''' Df, dFdp, dFds, dFdg, dFdsh = pyd.pydpsvfcj(len(xdata), xdata - pos, pos, sig, gam, shl) # Df,dFdp,dFds,dFdg,dFdsh = pyd.pydpsvfcjo(len(xdata),xdata-pos,pos,sig,gam,shl) return Df, dFdp, dFds, dFdg, dFdsh
[docs]def getPsVoigt(pos, sig, gam, xdata): 'needs a doc string' Df = pyd.pypsvoigt(len(xdata), xdata - pos, sig, gam) Df /= np.sum(Df) return Df
[docs]def getdPsVoigt(pos, sig, gam, xdata): 'needs a doc string' Df, dFdp, dFds, dFdg = pyd.pydpsvoigt(len(xdata), xdata - pos, sig, gam) return Df, dFdp, dFds, dFdg
[docs]def getEpsVoigt(pos, alp, bet, sig, gam, xdata): 'needs a doc string' Df = pyd.pyepsvoigt(len(xdata), xdata - pos, alp, bet, sig, gam) Df /= np.sum(Df) return Df
[docs]def getdEpsVoigt(pos, alp, bet, sig, gam, xdata): 'needs a doc string' Df, dFdp, dFda, dFdb, dFds, dFdg = pyd.pydepsvoigt(len(xdata), xdata - pos, alp, bet, sig, gam) return Df, dFdp, dFda, dFdb, dFds, dFdg
[docs]def ellipseSize(H, Sij, GB): 'Implements r=1/sqrt(sum((1/S)*(q.v)^2) per note from Alexander Brady' HX = np.inner(H.T, GB) lenHX = np.sqrt(np.sum(HX**2)) Esize, Rsize = nl.eigh(G2lat.U6toUij(Sij)) R = np.inner(HX / lenHX, Rsize)**2 * Esize #want column length for hkl in crystal lenR = 1. / np.sqrt(np.sum(R)) return lenR
[docs]def ellipseSizeDerv(H, Sij, GB): 'needs a doc string' lenR = ellipseSize(H, Sij, GB) delt = 0.001 dRdS = np.zeros(6) for i in range(6): Sij[i] -= delt lenM = ellipseSize(H, Sij, GB) Sij[i] += 2. * delt lenP = ellipseSize(H, Sij, GB) Sij[i] -= delt dRdS[i] = (lenP - lenM) / (2. * delt) return lenR, dRdS
[docs]def getHKLpeak(dmin, SGData, A, Inst=None, nodup=False): ''' Generates allowed by symmetry reflections with d >= dmin NB: GenHKLf & checkMagextc return True for extinct reflections :param dmin: minimum d-spacing :param SGData: space group data obtained from SpcGroup :param A: lattice parameter terms A1-A6 :param Inst: instrument parameter info :returns: HKLs: np.array hkl, etc for allowed reflections ''' HKL = G2lat.GenHLaue(dmin, SGData, A) HKLs = [] ds = [] for h, k, l, d in HKL: ext = G2spc.GenHKLf([h, k, l], SGData)[0] if ext and 'MagSpGrp' in SGData: ext = G2spc.checkMagextc([h, k, l], SGData) if not ext: if nodup and int(10000 * d) in ds: continue ds.append(int(10000 * d)) if Inst == None: HKLs.append([h, k, l, d, 0, -1]) else: HKLs.append([h, k, l, d, G2lat.Dsp2pos(Inst, d), -1]) return np.array(HKLs)
[docs]def getHKLMpeak(dmin, Inst, SGData, SSGData, Vec, maxH, A): 'needs a doc string' HKLs = [] vec = np.array(Vec) vstar = np.sqrt(G2lat.calc_rDsq( vec, A)) #find extra needed for -n SS reflections dvec = 1. / (maxH * vstar + 1. / dmin) HKL = G2lat.GenHLaue(dvec, SGData, A) SSdH = [vec * h for h in range(-maxH, maxH + 1)] SSdH = dict(zip(range(-maxH, maxH + 1), SSdH)) ifMag = False if 'MagSpGrp' in SGData: ifMag = True for h, k, l, d in HKL: ext = G2spc.GenHKLf([h, k, l], SGData)[0] if not ext and d >= dmin: HKLs.append([h, k, l, 0, d, G2lat.Dsp2pos(Inst, d), -1]) for dH in SSdH: if dH: DH = SSdH[dH] H = [h + DH[0], k + DH[1], l + DH[2]] d = float(1 / np.sqrt(G2lat.calc_rDsq(H, A))) if d >= dmin: HKLM = np.array([h, k, l, dH]) if G2spc.checkSSextc(HKLM, SSGData) or ifMag: HKLs.append( [h, k, l, dH, d, G2lat.Dsp2pos(Inst, d), -1]) return G2lat.sortHKLd(HKLs, True, True, True)
[docs]def getPeakProfile(dataType, parmDict, xdata, varyList, bakType): 'Computes the profile for a powder pattern' yb = getBackground('', parmDict, bakType, dataType, xdata)[0] yc = np.zeros_like(yb) cw = np.diff(xdata) cw = np.append(cw, cw[-1]) if 'C' in dataType: shl = max(parmDict['SH/L'], 0.002) Ka2 = False if 'Lam1' in parmDict.keys(): Ka2 = True lamRatio = 360 * (parmDict['Lam2'] - parmDict['Lam1']) / (np.pi * parmDict['Lam1']) kRatio = parmDict['I(L2)/I(L1)'] iPeak = 0 while True: try: pos = parmDict['pos' + str(iPeak)] tth = (pos - parmDict['Zero']) intens = parmDict['int' + str(iPeak)] sigName = 'sig' + str(iPeak) if sigName in varyList: sig = parmDict[sigName] else: sig = G2mth.getCWsig(parmDict, tth) sig = max(sig, 0.001) #avoid neg sigma^2 gamName = 'gam' + str(iPeak) if gamName in varyList: gam = parmDict[gamName] else: gam = G2mth.getCWgam(parmDict, tth) gam = max(gam, 0.001) #avoid neg gamma Wd, fmin, fmax = getWidthsCW(pos, sig, gam, shl) iBeg = np.searchsorted(xdata, pos - fmin) iFin = np.searchsorted(xdata, pos + fmin) if not iBeg + iFin: #peak below low limit iPeak += 1 continue elif not iBeg - iFin: #peak above high limit return yb + yc yc[iBeg:iFin] += intens * getFCJVoigt3(pos, sig, gam, shl, xdata[iBeg:iFin]) if Ka2: pos2 = pos + lamRatio * tand( pos / 2.0) # + 360/pi * Dlam/lam * tan(th) iBeg = np.searchsorted(xdata, pos2 - fmin) iFin = np.searchsorted(xdata, pos2 + fmin) if iBeg - iFin: yc[iBeg:iFin] += intens * kRatio * getFCJVoigt3( pos2, sig, gam, shl, xdata[iBeg:iFin]) iPeak += 1 except KeyError: #no more peaks to process return yb + yc else: Pdabc = parmDict['Pdabc'] difC = parmDict['difC'] iPeak = 0 while True: try: pos = parmDict['pos' + str(iPeak)] tof = pos - parmDict['Zero'] dsp = tof / difC intens = parmDict['int' + str(iPeak)] alpName = 'alp' + str(iPeak) if alpName in varyList: alp = parmDict[alpName] else: if len(Pdabc): alp = np.interp(dsp, Pdabc[0], Pdabc[1]) else: alp = G2mth.getTOFalpha(parmDict, dsp) alp = max(0.1, alp) betName = 'bet' + str(iPeak) if betName in varyList: bet = parmDict[betName] else: if len(Pdabc): bet = np.interp(dsp, Pdabc[0], Pdabc[2]) else: bet = G2mth.getTOFbeta(parmDict, dsp) bet = max(0.0001, bet) sigName = 'sig' + str(iPeak) if sigName in varyList: sig = parmDict[sigName] else: sig = G2mth.getTOFsig(parmDict, dsp) gamName = 'gam' + str(iPeak) if gamName in varyList: gam = parmDict[gamName] else: gam = G2mth.getTOFgamma(parmDict, dsp) gam = max(gam, 0.001) #avoid neg gamma Wd, fmin, fmax = getWidthsTOF(pos, alp, bet, sig, gam) iBeg = np.searchsorted(xdata, pos - fmin) iFin = np.searchsorted(xdata, pos + fmax) lenX = len(xdata) if not iBeg: iFin = np.searchsorted(xdata, pos + fmax) elif iBeg == lenX: iFin = iBeg else: iFin = np.searchsorted(xdata, pos + fmax) if not iBeg + iFin: #peak below low limit iPeak += 1 continue elif not iBeg - iFin: #peak above high limit return yb + yc yc[iBeg:iFin] += intens * getEpsVoigt(pos, alp, bet, sig, gam, xdata[iBeg:iFin]) iPeak += 1 except KeyError: #no more peaks to process return yb + yc
[docs]def getPeakProfileDerv(dataType, parmDict, xdata, varyList, bakType): 'needs a doc string' # needs to return np.array([dMdx1,dMdx2,...]) in same order as varylist = backVary,insVary,peakVary order dMdv = np.zeros(shape=(len(varyList), len(xdata))) dMdb, dMddb, dMdpk = getBackgroundDerv('', parmDict, bakType, dataType, xdata) if 'Back;0' in varyList: #background derivs are in front if present dMdv[0:len(dMdb)] = dMdb names = ['DebyeA', 'DebyeR', 'DebyeU'] for name in varyList: if 'Debye' in name: parm, Id = name.split(';') ip = names.index(parm) dMdv[varyList.index(name)] = dMddb[3 * int(Id) + ip] names = ['BkPkpos', 'BkPkint', 'BkPksig', 'BkPkgam'] for name in varyList: if 'BkPk' in name: parm, Id = name.split(';') ip = names.index(parm) dMdv[varyList.index(name)] = dMdpk[4 * int(Id) + ip] cw = np.diff(xdata) cw = np.append(cw, cw[-1]) if 'C' in dataType: shl = max(parmDict['SH/L'], 0.002) Ka2 = False if 'Lam1' in parmDict.keys(): Ka2 = True lamRatio = 360 * (parmDict['Lam2'] - parmDict['Lam1']) / (np.pi * parmDict['Lam1']) kRatio = parmDict['I(L2)/I(L1)'] iPeak = 0 while True: try: pos = parmDict['pos' + str(iPeak)] tth = (pos - parmDict['Zero']) intens = parmDict['int' + str(iPeak)] sigName = 'sig' + str(iPeak) if sigName in varyList: sig = parmDict[sigName] dsdU = dsdV = dsdW = 0 else: sig = G2mth.getCWsig(parmDict, tth) dsdU, dsdV, dsdW = G2mth.getCWsigDeriv(tth) sig = max(sig, 0.001) #avoid neg sigma gamName = 'gam' + str(iPeak) if gamName in varyList: gam = parmDict[gamName] dgdX = dgdY = dgdZ = 0 else: gam = G2mth.getCWgam(parmDict, tth) dgdX, dgdY, dgdZ = G2mth.getCWgamDeriv(tth) gam = max(gam, 0.001) #avoid neg gamma Wd, fmin, fmax = getWidthsCW(pos, sig, gam, shl) iBeg = np.searchsorted(xdata, pos - fmin) iFin = np.searchsorted(xdata, pos + fmin) if not iBeg + iFin: #peak below low limit iPeak += 1 continue elif not iBeg - iFin: #peak above high limit break dMdpk = np.zeros(shape=(6, len(xdata))) dMdipk = getdFCJVoigt3(pos, sig, gam, shl, xdata[iBeg:iFin]) for i in range(1, 5): dMdpk[i][ iBeg:iFin] += 100. * cw[iBeg:iFin] * intens * dMdipk[i] dMdpk[0][iBeg:iFin] += 100. * cw[iBeg:iFin] * dMdipk[0] dervDict = { 'int': dMdpk[0], 'pos': dMdpk[1], 'sig': dMdpk[2], 'gam': dMdpk[3], 'shl': dMdpk[4] } if Ka2: pos2 = pos + lamRatio * tand( pos / 2.0) # + 360/pi * Dlam/lam * tan(th) iBeg = np.searchsorted(xdata, pos2 - fmin) iFin = np.searchsorted(xdata, pos2 + fmin) if iBeg - iFin: dMdipk2 = getdFCJVoigt3(pos2, sig, gam, shl, xdata[iBeg:iFin]) for i in range(1, 5): dMdpk[i][iBeg:iFin] += 100. * cw[ iBeg:iFin] * intens * kRatio * dMdipk2[i] dMdpk[0][iBeg:iFin] += 100. * cw[ iBeg:iFin] * kRatio * dMdipk2[0] dMdpk[5][iBeg:iFin] += 100. * cw[iBeg:iFin] * dMdipk2[0] dervDict = { 'int': dMdpk[0], 'pos': dMdpk[1], 'sig': dMdpk[2], 'gam': dMdpk[3], 'shl': dMdpk[4], 'L1/L2': dMdpk[5] * intens } for parmName in ['pos', 'int', 'sig', 'gam']: try: idx = varyList.index(parmName + str(iPeak)) dMdv[idx] = dervDict[parmName] except ValueError: pass if 'U' in varyList: dMdv[varyList.index('U')] += dsdU * dervDict['sig'] if 'V' in varyList: dMdv[varyList.index('V')] += dsdV * dervDict['sig'] if 'W' in varyList: dMdv[varyList.index('W')] += dsdW * dervDict['sig'] if 'X' in varyList: dMdv[varyList.index('X')] += dgdX * dervDict['gam'] if 'Y' in varyList: dMdv[varyList.index('Y')] += dgdY * dervDict['gam'] if 'Z' in varyList: dMdv[varyList.index('Z')] += dgdZ * dervDict['gam'] if 'SH/L' in varyList: dMdv[varyList.index('SH/L')] += dervDict[ 'shl'] #problem here if 'I(L2)/I(L1)' in varyList: dMdv[varyList.index('I(L2)/I(L1)')] += dervDict['L1/L2'] iPeak += 1 except KeyError: #no more peaks to process break else: Pdabc = parmDict['Pdabc'] difC = parmDict['difC'] iPeak = 0 while True: try: pos = parmDict['pos' + str(iPeak)] tof = pos - parmDict['Zero'] dsp = tof / difC intens = parmDict['int' + str(iPeak)] alpName = 'alp' + str(iPeak) if alpName in varyList: alp = parmDict[alpName] else: if len(Pdabc): alp = np.interp(dsp, Pdabc[0], Pdabc[1]) dada0 = 0 else: alp = G2mth.getTOFalpha(parmDict, dsp) dada0 = G2mth.getTOFalphaDeriv(dsp) betName = 'bet' + str(iPeak) if betName in varyList: bet = parmDict[betName] else: if len(Pdabc): bet = np.interp(dsp, Pdabc[0], Pdabc[2]) dbdb0 = dbdb1 = dbdb2 = 0 else: bet = G2mth.getTOFbeta(parmDict, dsp) dbdb0, dbdb1, dbdb2 = G2mth.getTOFbetaDeriv(dsp) sigName = 'sig' + str(iPeak) if sigName in varyList: sig = parmDict[sigName] dsds0 = dsds1 = dsds2 = dsds3 = 0 else: sig = G2mth.getTOFsig(parmDict, dsp) dsds0, dsds1, dsds2, dsds3 = G2mth.getTOFsigDeriv(dsp) gamName = 'gam' + str(iPeak) if gamName in varyList: gam = parmDict[gamName] dsdX = dsdY = dsdZ = 0 else: gam = G2mth.getTOFgamma(parmDict, dsp) dsdX, dsdY, dsdZ = G2mth.getTOFgammaDeriv(dsp) gam = max(gam, 0.001) #avoid neg gamma Wd, fmin, fmax = getWidthsTOF(pos, alp, bet, sig, gam) iBeg = np.searchsorted(xdata, pos - fmin) lenX = len(xdata) if not iBeg: iFin = np.searchsorted(xdata, pos + fmax) elif iBeg == lenX: iFin = iBeg else: iFin = np.searchsorted(xdata, pos + fmax) if not iBeg + iFin: #peak below low limit iPeak += 1 continue elif not iBeg - iFin: #peak above high limit break dMdpk = np.zeros(shape=(7, len(xdata))) dMdipk = getdEpsVoigt(pos, alp, bet, sig, gam, xdata[iBeg:iFin]) for i in range(1, 6): dMdpk[i][iBeg:iFin] += intens * cw[iBeg:iFin] * dMdipk[i] dMdpk[0][iBeg:iFin] += cw[iBeg:iFin] * dMdipk[0] dervDict = { 'int': dMdpk[0], 'pos': dMdpk[1], 'alp': dMdpk[2], 'bet': dMdpk[3], 'sig': dMdpk[4], 'gam': dMdpk[5] } for parmName in ['pos', 'int', 'alp', 'bet', 'sig', 'gam']: try: idx = varyList.index(parmName + str(iPeak)) dMdv[idx] = dervDict[parmName] except ValueError: pass if 'alpha' in varyList: dMdv[varyList.index('alpha')] += dada0 * dervDict['alp'] if 'beta-0' in varyList: dMdv[varyList.index('beta-0')] += dbdb0 * dervDict['bet'] if 'beta-1' in varyList: dMdv[varyList.index('beta-1')] += dbdb1 * dervDict['bet'] if 'beta-q' in varyList: dMdv[varyList.index('beta-q')] += dbdb2 * dervDict['bet'] if 'sig-0' in varyList: dMdv[varyList.index('sig-0')] += dsds0 * dervDict['sig'] if 'sig-1' in varyList: dMdv[varyList.index('sig-1')] += dsds1 * dervDict['sig'] if 'sig-2' in varyList: dMdv[varyList.index('sig-2')] += dsds2 * dervDict['sig'] if 'sig-q' in varyList: dMdv[varyList.index('sig-q')] += dsds3 * dervDict['sig'] if 'X' in varyList: dMdv[varyList.index('X')] += dsdX * dervDict['gam'] if 'Y' in varyList: dMdv[varyList.index('Y')] += dsdY * dervDict['gam'] if 'Z' in varyList: dMdv[varyList.index('Z')] += dsdZ * dervDict['gam'] iPeak += 1 except KeyError: #no more peaks to process break return dMdv
[docs]def Dict2Values(parmdict, varylist): '''Use before call to leastsq to setup list of values for the parameters in parmdict, as selected by key in varylist''' return [parmdict[key] for key in varylist]
[docs]def Values2Dict(parmdict, varylist, values): ''' Use after call to leastsq to update the parameter dictionary with values corresponding to keys in varylist''' parmdict.update(zip(varylist, values))
[docs]def SetBackgroundParms(Background): 'Loads background parameters into dicts/lists to create varylist & parmdict' if len(Background) == 1: # fix up old backgrounds Background.append({'nDebye': 0, 'debyeTerms': []}) bakType, bakFlag = Background[0][:2] backVals = Background[0][3:] backNames = ['Back;' + str(i) for i in range(len(backVals))] Debye = Background[1] #also has background peaks stuff backDict = dict(zip(backNames, backVals)) backVary = [] if bakFlag: backVary = backNames backDict['nDebye'] = Debye['nDebye'] debyeDict = {} debyeList = [] for i in range(Debye['nDebye']): debyeNames = [ 'DebyeA;' + str(i), 'DebyeR;' + str(i), 'DebyeU;' + str(i) ] debyeDict.update(dict(zip(debyeNames, Debye['debyeTerms'][i][::2]))) debyeList += zip(debyeNames, Debye['debyeTerms'][i][1::2]) debyeVary = [] for item in debyeList: if item[1]: debyeVary.append(item[0]) backDict.update(debyeDict) backVary += debyeVary backDict['nPeaks'] = Debye['nPeaks'] peaksDict = {} peaksList = [] for i in range(Debye['nPeaks']): peaksNames = [ 'BkPkpos;' + str(i), 'BkPkint;' + str(i), 'BkPksig;' + str(i), 'BkPkgam;' + str(i) ] peaksDict.update(dict(zip(peaksNames, Debye['peaksList'][i][::2]))) peaksList += zip(peaksNames, Debye['peaksList'][i][1::2]) peaksVary = [] for item in peaksList: if item[1]: peaksVary.append(item[0]) backDict.update(peaksDict) backVary += peaksVary return bakType, backDict, backVary
[docs]def DoCalibInst(IndexPeaks, Inst): def SetInstParms(): dataType = Inst['Type'][0] insVary = [] insNames = [] insVals = [] for parm in Inst: insNames.append(parm) insVals.append(Inst[parm][1]) if parm in [ 'Lam', 'difC', 'difA', 'difB', 'Zero', ]: if Inst[parm][2]: insVary.append(parm) instDict = dict(zip(insNames, insVals)) return dataType, instDict, insVary def GetInstParms(parmDict, Inst, varyList): for name in Inst: Inst[name][1] = parmDict[name] def InstPrint(Inst, sigDict): print('Instrument Parameters:') if 'C' in Inst['Type'][0]: ptfmt = "%12.6f" else: ptfmt = "%12.3f" ptlbls = 'names :' ptstr = 'values:' sigstr = 'esds :' for parm in Inst: if parm in [ 'Lam', 'difC', 'difA', 'difB', 'Zero', ]: ptlbls += "%s" % (parm.center(12)) ptstr += ptfmt % (Inst[parm][1]) if parm in sigDict: sigstr += ptfmt % (sigDict[parm]) else: sigstr += 12 * ' ' print(ptlbls) print(ptstr) print(sigstr) def errPeakPos(values, peakDsp, peakPos, peakWt, dataType, parmDict, varyList): parmDict.update(zip(varyList, values)) return np.sqrt(peakWt) * ( G2lat.getPeakPos(dataType, parmDict, peakDsp) - peakPos) peakPos = [] peakDsp = [] peakWt = [] for peak, sig in zip(IndexPeaks[0], IndexPeaks[1]): if peak[2] and peak[3] and sig > 0.: peakPos.append(peak[0]) peakDsp.append(peak[-1]) #d-calc # peakWt.append(peak[-1]**2/sig**2) #weight by d**2 peakWt.append(1. / (sig * peak[-1])) # peakPos = np.array(peakPos) peakDsp = np.array(peakDsp) peakWt = np.array(peakWt) dataType, insDict, insVary = SetInstParms() parmDict = {} parmDict.update(insDict) varyList = insVary if not len(varyList): G2fil.G2Print('**** ERROR - nothing to refine! ****') return False while True: begin = time.time() values = np.array(Dict2Values(parmDict, varyList)) result = so.leastsq(errPeakPos, values, full_output=True, ftol=0.000001, args=(peakDsp, peakPos, peakWt, dataType, parmDict, varyList)) ncyc = int(result[2]['nfev'] / 2) runtime = time.time() - begin chisq = np.sum(result[2]['fvec']**2) Values2Dict(parmDict, varyList, result[0]) GOF = chisq / (len(peakPos) - len(varyList)) #reduced chi^2 G2fil.G2Print( 'Number of function calls: %d Number of observations: %d Number of parameters: %d' % (result[2]['nfev'], len(peakPos), len(varyList))) G2fil.G2Print('calib time = %8.3fs, %8.3fs/cycle' % (runtime, runtime / ncyc)) G2fil.G2Print('chi**2 = %12.6g, reduced chi**2 = %6.2f' % (chisq, GOF)) try: sig = np.sqrt(np.diag(result[1]) * GOF) if np.any(np.isnan(sig)): G2fil.G2Print( '*** Least squares aborted - some invalid esds possible ***' ) break #refinement succeeded - finish up! except ValueError: #result[1] is None on singular matrix G2fil.G2Print('**** Refinement failed - singular matrix ****') sigDict = dict(zip(varyList, sig)) GetInstParms(parmDict, Inst, varyList) InstPrint(Inst, sigDict) return True
[docs]def DoPeakFit(FitPgm, Peaks, Background, Limits, Inst, Inst2, data, fixback=None, prevVaryList=[], oneCycle=False, controls=None, dlg=None): '''Called to perform a peak fit, refining the selected items in the peak table as well as selected items in the background. :param str FitPgm: type of fit to perform. At present this is ignored. :param list Peaks: a list of peaks. Each peak entry is a list with 8 values: four values followed by a refine flag where the values are: position, intensity, sigma (Gaussian width) and gamma (Lorentzian width). From the Histogram/"Peak List" tree entry, dict item "peaks" :param list Background: describes the background. List with two items. Item 0 specifies a background model and coefficients. Item 1 is a dict. From the Histogram/Background tree entry. :param list Limits: min and max x-value to use :param dict Inst: Instrument parameters :param dict Inst2: more Instrument parameters :param numpy.array data: a 5xn array. data[0] is the x-values, data[1] is the y-values, data[2] are weight values, data[3], [4] and [5] are calc, background and difference intensities, respectively. :param array fixback: fixed background values :param list prevVaryList: Used in sequential refinements to override the variable list. Defaults as an empty list. :param bool oneCycle: True if only one cycle of fitting should be performed :param dict controls: a dict specifying two values, Ftol = controls['min dM/M'] and derivType = controls['deriv type']. If None default values are used. :param wx.Dialog dlg: A dialog box that is updated with progress from the fit. Defaults to None, which means no updates are done. ''' def GetBackgroundParms(parmList, Background): iBak = 0 while True: try: bakName = 'Back;' + str(iBak) Background[0][iBak + 3] = parmList[bakName] iBak += 1 except KeyError: break iDb = 0 while True: names = ['DebyeA;', 'DebyeR;', 'DebyeU;'] try: for i, name in enumerate(names): val = parmList[name + str(iDb)] Background[1]['debyeTerms'][iDb][2 * i] = val iDb += 1 except KeyError: break iDb = 0 while True: names = ['BkPkpos;', 'BkPkint;', 'BkPksig;', 'BkPkgam;'] try: for i, name in enumerate(names): val = parmList[name + str(iDb)] Background[1]['peaksList'][iDb][2 * i] = val iDb += 1 except KeyError: break def BackgroundPrint(Background, sigDict): print('Background coefficients for ' + Background[0][0] + ' function') ptfmt = "%12.5f" ptstr = 'value: ' sigstr = 'esd : ' for i, back in enumerate(Background[0][3:]): ptstr += ptfmt % (back) if Background[0][1]: prm = 'Back;' + str(i) if prm in sigDict: sigstr += ptfmt % (sigDict[prm]) else: sigstr += " " * 12 if len(ptstr) > 75: print(ptstr) if Background[0][1]: print(sigstr) ptstr = 'value: ' sigstr = 'esd : ' if len(ptstr) > 8: print(ptstr) if Background[0][1]: print(sigstr) if Background[1]['nDebye']: parms = ['DebyeA;', 'DebyeR;', 'DebyeU;'] print('Debye diffuse scattering coefficients') ptfmt = "%12.5f" print( ' term DebyeA esd DebyeR esd DebyeU esd' ) for term in range(Background[1]['nDebye']): line = ' term %d' % (term) for ip, name in enumerate(parms): line += ptfmt % (Background[1]['debyeTerms'][term][2 * ip]) if name + str(term) in sigDict: line += ptfmt % (sigDict[name + str(term)]) else: line += " " * 12 print(line) if Background[1]['nPeaks']: print('Coefficients for Background Peaks') ptfmt = "%15.3f" for j, pl in enumerate(Background[1]['peaksList']): names = 'peak %3d:' % (j + 1) ptstr = 'values :' sigstr = 'esds :' for i, lbl in enumerate( ['BkPkpos', 'BkPkint', 'BkPksig', 'BkPkgam']): val = pl[2 * i] prm = lbl + ";" + str(j) names += '%15s' % (prm) ptstr += ptfmt % (val) if prm in sigDict: sigstr += ptfmt % (sigDict[prm]) else: sigstr += " " * 15 print(names) print(ptstr) print(sigstr) def SetInstParms(Inst): dataType = Inst['Type'][0] insVary = [] insNames = [] insVals = [] for parm in Inst: insNames.append(parm) insVals.append(Inst[parm][1]) if parm in [ 'U', 'V', 'W', 'X', 'Y', 'Z', 'SH/L', 'I(L2)/I(L1)', 'alpha', 'beta-0', 'beta-1', 'beta-q', 'sig-0', 'sig-1', 'sig-2', 'sig-q', ] and Inst[parm][2]: insVary.append(parm) instDict = dict(zip(insNames, insVals)) # instDict['X'] = max(instDict['X'],0.01) # instDict['Y'] = max(instDict['Y'],0.01) if 'SH/L' in instDict: instDict['SH/L'] = max(instDict['SH/L'], 0.002) return dataType, instDict, insVary def GetInstParms(parmDict, Inst, varyList, Peaks): for name in Inst: Inst[name][1] = parmDict[name] iPeak = 0 while True: try: sigName = 'sig' + str(iPeak) pos = parmDict['pos' + str(iPeak)] if sigName not in varyList: if 'C' in Inst['Type'][0]: parmDict[sigName] = G2mth.getCWsig(parmDict, pos) else: dsp = G2lat.Pos2dsp(Inst, pos) parmDict[sigName] = G2mth.getTOFsig(parmDict, dsp) gamName = 'gam' + str(iPeak) if gamName not in varyList: if 'C' in Inst['Type'][0]: parmDict[gamName] = G2mth.getCWgam(parmDict, pos) else: dsp = G2lat.Pos2dsp(Inst, pos) parmDict[gamName] = G2mth.getTOFgamma(parmDict, dsp) iPeak += 1 except KeyError: break def InstPrint(Inst, sigDict): print('Instrument Parameters:') ptfmt = "%12.6f" ptlbls = 'names :' ptstr = 'values:' sigstr = 'esds :' for parm in Inst: if parm in [ 'U', 'V', 'W', 'X', 'Y', 'Z', 'SH/L', 'I(L2)/I(L1)', 'alpha', 'beta-0', 'beta-1', 'beta-q', 'sig-0', 'sig-1', 'sig-2', 'sig-q', ]: ptlbls += "%s" % (parm.center(12)) ptstr += ptfmt % (Inst[parm][1]) if parm in sigDict: sigstr += ptfmt % (sigDict[parm]) else: sigstr += 12 * ' ' print(ptlbls) print(ptstr) print(sigstr) def SetPeaksParms(dataType, Peaks): peakNames = [] peakVary = [] peakVals = [] if 'C' in dataType: names = ['pos', 'int', 'sig', 'gam'] else: names = ['pos', 'int', 'alp', 'bet', 'sig', 'gam'] for i, peak in enumerate(Peaks): for j, name in enumerate(names): peakVals.append(peak[2 * j]) parName = name + str(i) peakNames.append(parName) if peak[2 * j + 1]: peakVary.append(parName) return dict(zip(peakNames, peakVals)), peakVary def GetPeaksParms(Inst, parmDict, Peaks, varyList): if 'C' in Inst['Type'][0]: names = ['pos', 'int', 'sig', 'gam'] else: #'T' names = ['pos', 'int', 'alp', 'bet', 'sig', 'gam'] for i, peak in enumerate(Peaks): pos = parmDict['pos' + str(i)] if 'difC' in Inst: dsp = pos / Inst['difC'][1] for j in range(len(names)): parName = names[j] + str(i) if parName in varyList: peak[2 * j] = parmDict[parName] elif 'alpha' in parName: peak[2 * j] = parmDict['alpha'] / dsp elif 'beta' in parName: peak[2 * j] = G2mth.getTOFbeta(parmDict, dsp) elif 'sig' in parName: if 'C' in Inst['Type'][0]: peak[2 * j] = G2mth.getCWsig(parmDict, pos) else: peak[2 * j] = G2mth.getTOFsig(parmDict, dsp) elif 'gam' in parName: if 'C' in Inst['Type'][0]: peak[2 * j] = G2mth.getCWgam(parmDict, pos) else: peak[2 * j] = G2mth.getTOFgamma(parmDict, dsp) def PeaksPrint(dataType, parmDict, sigDict, varyList, ptsperFW): print('Peak coefficients:') if 'C' in dataType: names = ['pos', 'int', 'sig', 'gam'] else: #'T' names = ['pos', 'int', 'alp', 'bet', 'sig', 'gam'] head = 13 * ' ' for name in names: if name in ['alp', 'bet']: head += name.center(8) + 'esd'.center(8) else: head += name.center(10) + 'esd'.center(10) head += 'bins'.center(8) print(head) if 'C' in dataType: ptfmt = { 'pos': "%10.5f", 'int': "%10.1f", 'sig': "%10.3f", 'gam': "%10.3f" } else: ptfmt = { 'pos': "%10.2f", 'int': "%10.4f", 'alp': "%8.3f", 'bet': "%8.5f", 'sig': "%10.3f", 'gam': "%10.3f" } for i, peak in enumerate(Peaks): ptstr = ':' for j in range(len(names)): name = names[j] parName = name + str(i) ptstr += ptfmt[name] % (parmDict[parName]) if parName in varyList: ptstr += ptfmt[name] % (sigDict[parName]) else: if name in ['alp', 'bet']: ptstr += 8 * ' ' else: ptstr += 10 * ' ' ptstr += '%9.2f' % (ptsperFW[i]) print('%s' % (('Peak' + str(i + 1)).center(8)), ptstr) def devPeakProfile(values, xdata, ydata, weights, dataType, parmdict, varylist, bakType, dlg): parmdict.update(zip(varylist, values)) return np.sqrt(weights) * getPeakProfileDerv(dataType, parmdict, xdata, varylist, bakType) def errPeakProfile(values, xdata, ydata, weights, dataType, parmdict, varylist, bakType, dlg): parmdict.update(zip(varylist, values)) M = np.sqrt(weights) * (getPeakProfile(dataType, parmdict, xdata, varylist, bakType) - ydata) Rwp = min(100., np.sqrt(np.sum(M**2) / np.sum(weights * ydata**2)) * 100.) if dlg: GoOn = dlg.Update(Rwp, newmsg='%s%8.3f%s' % ('Peak fit Rwp =', Rwp, '%'))[0] if not GoOn: return -M #abort!! return M if controls: Ftol = controls['min dM/M'] else: Ftol = 0.0001 if oneCycle: Ftol = 1.0 x, y, w, yc, yb, yd = data #these are numpy arrays - remove masks! if fixback is None: fixback = np.zeros_like(y) yc *= 0. #set calcd ones to zero yb *= 0. yd *= 0. cw = x[1:] - x[:-1] xBeg = np.searchsorted(x, Limits[0]) xFin = np.searchsorted(x, Limits[1]) + 1 bakType, bakDict, bakVary = SetBackgroundParms(Background) dataType, insDict, insVary = SetInstParms(Inst) peakDict, peakVary = SetPeaksParms(Inst['Type'][0], Peaks) parmDict = {} parmDict.update(bakDict) parmDict.update(insDict) parmDict.update(peakDict) parmDict['Pdabc'] = [] #dummy Pdabc parmDict.update(Inst2) #put in real one if there if prevVaryList: varyList = prevVaryList[:] else: varyList = bakVary + insVary + peakVary fullvaryList = varyList[:] while True: begin = time.time() values = np.array(Dict2Values(parmDict, varyList)) Rvals = {} badVary = [] result = so.leastsq(errPeakProfile, values, Dfun=devPeakProfile, full_output=True, ftol=Ftol, col_deriv=True, args=(x[xBeg:xFin], (y + fixback)[xBeg:xFin], w[xBeg:xFin], dataType, parmDict, varyList, bakType, dlg)) ncyc = int(result[2]['nfev'] / 2) runtime = time.time() - begin chisq = np.sum(result[2]['fvec']**2) Values2Dict(parmDict, varyList, result[0]) Rvals['Rwp'] = np.sqrt( chisq / np.sum(w[xBeg:xFin] * (y + fixback)[xBeg:xFin]**2)) * 100. #to % Rvals['GOF'] = chisq / (xFin - xBeg - len(varyList)) #reduced chi^2 G2fil.G2Print( 'Number of function calls: %d Number of observations: %d Number of parameters: %d' % (result[2]['nfev'], xFin - xBeg, len(varyList))) if ncyc: G2fil.G2Print('fitpeak time = %8.3fs, %8.3fs/cycle' % (runtime, runtime / ncyc)) G2fil.G2Print('Rwp = %7.2f%%, chi**2 = %12.6g, reduced chi**2 = %6.2f' % (Rvals['Rwp'], chisq, Rvals['GOF'])) sig = [0] * len(varyList) if len(varyList) == 0: break # if nothing was refined try: sig = np.sqrt(np.diag(result[1]) * Rvals['GOF']) if np.any(np.isnan(sig)): G2fil.G2Print( '*** Least squares aborted - some invalid esds possible ***' ) break #refinement succeeded - finish up! except ValueError: #result[1] is None on singular matrix G2fil.G2Print('**** Refinement failed - singular matrix ****') Ipvt = result[2]['ipvt'] for i, ipvt in enumerate(Ipvt): if not np.sum(result[2]['fjac'], axis=1)[i]: G2fil.G2Print('Removing parameter: ' + varyList[ipvt - 1]) badVary.append(varyList[ipvt - 1]) del (varyList[ipvt - 1]) break else: # nothing removed break if dlg: dlg.Destroy() sigDict = dict(zip(varyList, sig)) yb[xBeg:xFin] = getBackground('', parmDict, bakType, dataType, x[xBeg:xFin])[0] - fixback[xBeg:xFin] yc[xBeg:xFin] = getPeakProfile(dataType, parmDict, x[xBeg:xFin], varyList, bakType) - fixback[xBeg:xFin] yd[xBeg:xFin] = y[xBeg:xFin] - yc[xBeg:xFin] GetBackgroundParms(parmDict, Background) if bakVary: BackgroundPrint(Background, sigDict) GetInstParms(parmDict, Inst, varyList, Peaks) if insVary: InstPrint(Inst, sigDict) GetPeaksParms(Inst, parmDict, Peaks, varyList) binsperFWHM = [] for peak in Peaks: FWHM = getFWHM(peak[0], Inst) try: binsperFWHM.append(FWHM / cw[x.searchsorted(peak[0])]) except IndexError: binsperFWHM.append(0.) if peakVary: PeaksPrint(dataType, parmDict, sigDict, varyList, binsperFWHM) if len(binsperFWHM): if min(binsperFWHM) < 1.: G2fil.G2Print( '*** Warning: calculated peak widths are too narrow to refine profile coefficients ***' ) if 'T' in Inst['Type'][0]: G2fil.G2Print( ' Manually increase sig-0, 1, or 2 in Instrument Parameters' ) else: G2fil.G2Print(' Manually increase W in Instrument Parameters') elif min(binsperFWHM) < 4.: G2fil.G2Print( '*** Warning: data binning yields too few data points across peak FWHM for reliable Rietveld refinement ***' ) G2fil.G2Print('*** recommended is 6-10; you have %.2f ***' % (min(binsperFWHM))) return sigDict, result, sig, Rvals, varyList, parmDict, fullvaryList, badVary
[docs]def calcIncident(Iparm, xdata): 'needs a doc string' def IfunAdv(Iparm, xdata): Itype = Iparm['Itype'] Icoef = Iparm['Icoeff'] DYI = np.ones((12, xdata.shape[0])) YI = np.ones_like(xdata) * Icoef[0] x = xdata / 1000. #expressions are in ms if Itype == 'Exponential': for i in [1, 3, 5, 7, 9]: Eterm = np.exp(-Icoef[i + 1] * x**((i + 1) / 2)) YI += Icoef[i] * Eterm DYI[i] *= Eterm DYI[i + 1] *= -Icoef[i] * Eterm * x**((i + 1) / 2) elif 'Maxwell' in Itype: Eterm = np.exp(-Icoef[2] / x**2) DYI[1] = Eterm / x**5 DYI[2] = -Icoef[1] * DYI[1] / x**2 YI += (Icoef[1] * Eterm / x**5) if 'Exponential' in Itype: for i in range(3, 11, 2): Eterm = np.exp(-Icoef[i + 1] * x**((i + 1) / 2)) YI += Icoef[i] * Eterm DYI[i] *= Eterm DYI[i + 1] *= -Icoef[i] * Eterm * x**((i + 1) / 2) else: #Chebyschev T = (2. / x) - 1. Ccof = np.ones((12, xdata.shape[0])) Ccof[1] = T for i in range(2, 12): Ccof[i] = 2 * T * Ccof[i - 1] - Ccof[i - 2] for i in range(1, 10): YI += Ccof[i] * Icoef[i + 2] DYI[i + 2] = Ccof[i] return YI, DYI Iesd = np.array(Iparm['Iesd']) Icovar = Iparm['Icovar'] YI, DYI = IfunAdv(Iparm, xdata) YI = np.where(YI > 0, YI, 1.) WYI = np.zeros_like(xdata) vcov = np.zeros((12, 12)) k = 0 for i in range(12): for j in range(i, 12): vcov[i][j] = Icovar[k] * Iesd[i] * Iesd[j] vcov[j][i] = Icovar[k] * Iesd[i] * Iesd[j] k += 1 M = np.inner(vcov, DYI.T) WYI = np.sum(M * DYI, axis=0) WYI = np.where(WYI > 0., WYI, 0.) return YI, WYI
################################################################################ # Reflectometry calculations ################################################################################
[docs]def REFDRefine(Profile, ProfDict, Inst, Limits, Substances, data): G2fil.G2Print('fit REFD data by ' + data['Minimizer'] + ' using %.2f%% data resolution' % (data['Resolution'][0])) class RandomDisplacementBounds(object): """random displacement with bounds""" def __init__(self, xmin, xmax, stepsize=0.5): self.xmin = xmin self.xmax = xmax self.stepsize = stepsize def __call__(self, x): """take a random step but ensure the new position is within the bounds""" while True: # this could be done in a much more clever way, but it will work for example purposes steps = self.xmax - self.xmin xnew = x + np.random.uniform(-self.stepsize * steps, self.stepsize * steps, np.shape(x)) if np.all(xnew < self.xmax) and np.all(xnew > self.xmin): break return xnew def GetModelParms(): parmDict = {} varyList = [] values = [] bounds = [] parmDict['dQ type'] = data['dQ type'] parmDict['Res'] = data['Resolution'][0] / (100. * sateln2 ) #% FWHM-->decimal sig for parm in ['Scale', 'FltBack']: parmDict[parm] = data[parm][0] if data[parm][1]: varyList.append(parm) values.append(data[parm][0]) bounds.append(Bounds[parm]) parmDict['Layer Seq'] = np.array([ '0', ] + data['Layer Seq'].split() + [ str(len(data['Layers']) - 1), ], dtype=int) parmDict['nLayers'] = len(parmDict['Layer Seq']) for ilay, layer in enumerate(data['Layers']): name = layer['Name'] cid = str(ilay) + ';' parmDict[cid + 'Name'] = name for parm in ['Thick', 'Rough', 'DenMul', 'Mag SLD', 'iDenMul']: parmDict[cid + parm] = layer.get(parm, [0., False])[0] if layer.get(parm, [0., False])[1]: varyList.append(cid + parm) value = layer[parm][0] values.append(value) if value: bound = [value * Bfac, value / Bfac] else: bound = [0., 10.] bounds.append(bound) if name not in ['vacuum', 'unit scatter']: parmDict[cid + 'rho'] = Substances[name]['Scatt density'] parmDict[cid + 'irho'] = Substances[name].get( 'XImag density', 0.) return parmDict, varyList, values, bounds def SetModelParms(): line = ' Refined parameters: Histogram scale: %.4g' % ( parmDict['Scale']) if 'Scale' in varyList: data['Scale'][0] = parmDict['Scale'] line += ' esd: %.4g' % (sigDict['Scale']) G2fil.G2Print(line) line = ' Flat background: %15.4g' % (parmDict['FltBack']) if 'FltBack' in varyList: data['FltBack'][0] = parmDict['FltBack'] line += ' esd: %15.3g' % (sigDict['FltBack']) G2fil.G2Print(line) for ilay, layer in enumerate(data['Layers']): name = layer['Name'] G2fil.G2Print(' Parameters for layer: %d %s' % (ilay, name)) cid = str(ilay) + ';' line = ' ' line2 = ' Scattering density: Real %.5g' % ( Substances[name]['Scatt density'] * parmDict[cid + 'DenMul']) line2 += ' Imag %.5g' % (Substances[name].get('XImag density', 0.) * parmDict[cid + 'DenMul']) for parm in ['Thick', 'Rough', 'DenMul', 'Mag SLD', 'iDenMul']: if parm in layer: if parm == 'Rough': layer[parm][0] = abs(parmDict[cid + parm]) #make positive else: layer[parm][0] = parmDict[cid + parm] line += ' %s: %.3f' % (parm, layer[parm][0]) if cid + parm in varyList: line += ' esd: %.3g' % (sigDict[cid + parm]) G2fil.G2Print(line) G2fil.G2Print(line2) def calcREFD(values, Q, Io, wt, Qsig, parmDict, varyList): parmDict.update(zip(varyList, values)) M = np.sqrt(wt) * (getREFD(Q, Qsig, parmDict) - Io) return M def sumREFD(values, Q, Io, wt, Qsig, parmDict, varyList): parmDict.update(zip(varyList, values)) M = np.sqrt(wt) * (getREFD(Q, Qsig, parmDict) - Io) return np.sum(M**2) def getREFD(Q, Qsig, parmDict): Ic = np.ones_like(Q) * parmDict['FltBack'] Scale = parmDict['Scale'] Nlayers = parmDict['nLayers'] Res = parmDict['Res'] depth = np.zeros(Nlayers) rho = np.zeros(Nlayers) irho = np.zeros(Nlayers) sigma = np.zeros(Nlayers) for ilay, lay in enumerate(parmDict['Layer Seq']): cid = str(lay) + ';' depth[ilay] = parmDict[cid + 'Thick'] sigma[ilay] = parmDict[cid + 'Rough'] if parmDict[cid + 'Name'] == u'unit scatter': rho[ilay] = parmDict[cid + 'DenMul'] irho[ilay] = parmDict[cid + 'iDenMul'] elif 'vacuum' != parmDict[cid + 'Name']: rho[ilay] = parmDict[cid + 'rho'] * parmDict[cid + 'DenMul'] irho[ilay] = parmDict[cid + 'irho'] * parmDict[cid + 'DenMul'] if cid + 'Mag SLD' in parmDict: rho[ilay] += parmDict[cid + 'Mag SLD'] if parmDict['dQ type'] == 'None': AB = abeles(0.5 * Q, depth, rho, irho, sigma[1:]) #Q --> k, offset roughness for abeles elif 'const' in parmDict['dQ type']: AB = SmearAbeles(0.5 * Q, Q * Res, depth, rho, irho, sigma[1:]) else: #dQ/Q in data AB = SmearAbeles(0.5 * Q, Qsig, depth, rho, irho, sigma[1:]) Ic += AB * Scale return Ic def estimateT0(takestep): Mmax = -1.e-10 Mmin = 1.e10 for i in range(100): x0 = takestep(values) M = sumREFD(x0, Q[Ibeg:Ifin], Io[Ibeg:Ifin], wtFactor * wt[Ibeg:Ifin], Qsig[Ibeg:Ifin], parmDict, varyList) Mmin = min(M, Mmin) MMax = max(M, Mmax) return 1.5 * (MMax - Mmin) Q, Io, wt, Ic, Ib, Qsig = Profile[:6] if data.get('2% weight'): wt = 1. / (0.02 * Io)**2 Qmin = Limits[1][0] Qmax = Limits[1][1] wtFactor = ProfDict['wtFactor'] Bfac = data['Toler'] Ibeg = np.searchsorted(Q, Qmin) Ifin = np.searchsorted(Q, Qmax) + 1 #include last point Ic[:] = 0 Bounds = { 'Scale': [data['Scale'][0] * Bfac, data['Scale'][0] / Bfac], 'FltBack': [0., 1.e-6], 'DenMul': [0., 1.], 'Thick': [1., 500.], 'Rough': [0., 10.], 'Mag SLD': [-10., 10.], 'iDenMul': [-1., 1.] } parmDict, varyList, values, bounds = GetModelParms() Msg = 'Failed to converge' if varyList: if data['Minimizer'] == 'LMLS': result = so.leastsq(calcREFD, values, full_output=True, epsfcn=1.e-8, ftol=1.e-6, args=(Q[Ibeg:Ifin], Io[Ibeg:Ifin], wtFactor * wt[Ibeg:Ifin], Qsig[Ibeg:Ifin], parmDict, varyList)) parmDict.update(zip(varyList, result[0])) chisq = np.sum(result[2]['fvec']**2) ncalc = result[2]['nfev'] covM = result[1] newVals = result[0] elif data['Minimizer'] == 'Basin Hopping': xyrng = np.array(bounds).T take_step = RandomDisplacementBounds(xyrng[0], xyrng[1]) T0 = estimateT0(take_step) G2fil.G2Print(' Estimated temperature: %.3g' % (T0)) result = so.basinhopping(sumREFD, values, take_step=take_step, disp=True, T=T0, stepsize=Bfac, interval=20, niter=200, minimizer_kwargs={ 'method': 'L-BFGS-B', 'bounds': bounds, 'args': (Q[Ibeg:Ifin], Io[Ibeg:Ifin], wtFactor * wt[Ibeg:Ifin], Qsig[Ibeg:Ifin], parmDict, varyList) }) chisq = result.fun ncalc = result.nfev newVals = result.x covM = [] elif data['Minimizer'] == 'MC/SA Anneal': xyrng = np.array(bounds).T result = G2mth.anneal(sumREFD, values, args=(Q[Ibeg:Ifin], Io[Ibeg:Ifin], wtFactor * wt[Ibeg:Ifin], Qsig[Ibeg:Ifin], parmDict, varyList), schedule='log', full_output=True, maxeval=None, maxaccept=None, maxiter=10, dwell=1000, boltzmann=10.0, feps=1e-6, lower=xyrng[0], upper=xyrng[1], slope=0.9, ranStart=True, ranRange=0.20, autoRan=False, dlg=None) newVals = result[0] parmDict.update(zip(varyList, newVals)) chisq = result[1] ncalc = result[3] covM = [] G2fil.G2Print(' MC/SA final temperature: %.4g' % (result[2])) elif data['Minimizer'] == 'L-BFGS-B': result = so.minimize( sumREFD, values, method='L-BFGS-B', bounds=bounds, #ftol=Ftol, args=(Q[Ibeg:Ifin], Io[Ibeg:Ifin], wtFactor * wt[Ibeg:Ifin], Qsig[Ibeg:Ifin], parmDict, varyList)) parmDict.update(zip(varyList, result['x'])) chisq = result.fun ncalc = result.nfev newVals = result.x covM = [] else: #nothing varied M = calcREFD(values, Q[Ibeg:Ifin], Io[Ibeg:Ifin], wtFactor * wt[Ibeg:Ifin], Qsig[Ibeg:Ifin], parmDict, varyList) chisq = np.sum(M**2) ncalc = 0 covM = [] sig = [] sigDict = {} result = [] Rvals = {} Rvals['Rwp'] = np.sqrt( chisq / np.sum(wt[Ibeg:Ifin] * Io[Ibeg:Ifin]**2)) * 100. #to % Rvals['GOF'] = chisq / (Ifin - Ibeg - len(varyList)) #reduced chi^2 Ic[Ibeg:Ifin] = getREFD(Q[Ibeg:Ifin], Qsig[Ibeg:Ifin], parmDict) Ib[Ibeg:Ifin] = parmDict['FltBack'] try: if not len(varyList): Msg += ' - nothing refined' raise ValueError Nans = np.isnan(newVals) if np.any(Nans): name = varyList[Nans.nonzero(True)[0]] Msg += ' Nan result for ' + name + '!' raise ValueError Negs = np.less_equal(newVals, 0.) if np.any(Negs): indx = Negs.nonzero() name = varyList[indx[0][0]] if name != 'FltBack' and name.split(';')[1] in [ 'Thick', ]: Msg += ' negative coefficient for ' + name + '!' raise ValueError if len(covM): sig = np.sqrt(np.diag(covM) * Rvals['GOF']) covMatrix = covM * Rvals['GOF'] else: sig = np.zeros(len(varyList)) covMatrix = [] sigDict = dict(zip(varyList, sig)) G2fil.G2Print(' Results of reflectometry data modelling fit:') G2fil.G2Print( 'Number of function calls: %d Number of observations: %d Number of parameters: %d' % (ncalc, Ifin - Ibeg, len(varyList))) G2fil.G2Print('Rwp = %7.2f%%, chi**2 = %12.6g, reduced chi**2 = %6.2f' % (Rvals['Rwp'], chisq, Rvals['GOF'])) SetModelParms() return True, result, varyList, sig, Rvals, covMatrix, parmDict, '' except (ValueError, TypeError): #when bad LS refinement; covM missing or with nans G2fil.G2Print(Msg) return False, 0, 0, 0, 0, 0, 0, Msg
[docs]def makeSLDprofile(data, Substances): sq2 = np.sqrt(2.) laySeq = [ '0', ] + data['Layer Seq'].split() + [ str(len(data['Layers']) - 1), ] Nlayers = len(laySeq) laySeq = np.array(laySeq, dtype=int) interfaces = np.zeros(Nlayers) rho = np.zeros(Nlayers) sigma = np.zeros(Nlayers) name = data['Layers'][0]['Name'] thick = 0. for ilay, lay in enumerate(laySeq): layer = data['Layers'][lay] name = layer['Name'] if 'Thick' in layer: thick += layer['Thick'][0] interfaces[ilay] = layer['Thick'][0] + interfaces[ilay - 1] if 'Rough' in layer: sigma[ilay] = max(0.001, layer['Rough'][0]) if name != 'vacuum': if name == 'unit scatter': rho[ilay] = np.sqrt(layer['DenMul'][0]**2 + layer['iDenMul'][0]**2) else: rrho = Substances[name]['Scatt density'] irho = Substances[name]['XImag density'] rho[ilay] = np.sqrt(rrho**2 + irho**2) * layer['DenMul'][0] if 'Mag SLD' in layer: rho[ilay] += layer['Mag SLD'][0] name = data['Layers'][-1]['Name'] x = np.linspace(-0.15 * thick, 1.15 * thick, 1000, endpoint=True) xr = np.flipud(x) interfaces[-1] = x[-1] y = np.ones_like(x) * rho[0] iBeg = 0 for ilayer in range(Nlayers - 1): delt = rho[ilayer + 1] - rho[ilayer] iPos = np.searchsorted(x, interfaces[ilayer]) y[iBeg:] += (delt / 2.) * sp.erfc( (interfaces[ilayer] - x[iBeg:]) / (sq2 * sigma[ilayer + 1])) iBeg = iPos return x, xr, y
[docs]def REFDModelFxn(Profile, Inst, Limits, Substances, data): Q, Io, wt, Ic, Ib, Qsig = Profile[:6] Qmin = Limits[1][0] Qmax = Limits[1][1] iBeg = np.searchsorted(Q, Qmin) iFin = np.searchsorted(Q, Qmax) + 1 #include last point Ib[:] = data['FltBack'][0] Ic[:] = 0 Scale = data['Scale'][0] if data['Layer Seq'] == []: return laySeq = [ '0', ] + data['Layer Seq'].split() + [ str(len(data['Layers']) - 1), ] Nlayers = len(laySeq) depth = np.zeros(Nlayers) rho = np.zeros(Nlayers) irho = np.zeros(Nlayers) sigma = np.zeros(Nlayers) for ilay, lay in enumerate(np.array(laySeq, dtype=int)): layer = data['Layers'][lay] name = layer['Name'] if 'Thick' in layer: #skips first & last layers depth[ilay] = layer['Thick'][0] if 'Rough' in layer: #skips first layer sigma[ilay] = layer['Rough'][0] if 'unit scatter' == name: rho[ilay] = layer['DenMul'][0] irho[ilay] = layer['iDenMul'][0] else: rho[ilay] = Substances[name]['Scatt density'] * layer['DenMul'][0] irho[ilay] = Substances[name].get('XImag density', 0.) * layer['DenMul'][0] if 'Mag SLD' in layer: rho[ilay] += layer['Mag SLD'][0] if data['dQ type'] == 'None': AB = abeles(0.5 * Q[iBeg:iFin], depth, rho, irho, sigma[1:]) #Q --> k, offset roughness for abeles elif 'const' in data['dQ type']: res = data['Resolution'][0] / (100. * sateln2) AB = SmearAbeles(0.5 * Q[iBeg:iFin], res * Q[iBeg:iFin], depth, rho, irho, sigma[1:]) else: #dQ/Q in data AB = SmearAbeles(0.5 * Q[iBeg:iFin], Qsig[iBeg:iFin], depth, rho, irho, sigma[1:]) Ic[iBeg:iFin] = AB * Scale + Ib[iBeg:iFin]
[docs]def abeles(kz, depth, rho, irho=0, sigma=0): """ Optical matrix form of the reflectivity calculation. O.S. Heavens, Optical Properties of Thin Solid Films Reflectometry as a function of kz for a set of slabs. :param kz: float[n] (1/Ang). Scattering vector, :math:`2pi sin(theta)/lambda`. This is :math:`\\tfrac12 Q_z`. :param depth: float[m] (Ang). thickness of each layer. The thickness of the incident medium and substrate are ignored. :param rho: float[n,k] (1e-6/Ang^2) Real scattering length density for each layer for each kz :param irho: float[n,k] (1e-6/Ang^2) Imaginary scattering length density for each layer for each kz Note: absorption cross section mu = 2 irho/lambda for neutrons :param sigma: float[m-1] (Ang) interfacial roughness. This is the roughness between a layer and the previous layer. The sigma array should have m-1 entries. Slabs are ordered with the surface SLD at index 0 and substrate at index -1, or reversed if kz < 0. """ def calc(kz, depth, rho, irho, sigma): if len(kz) == 0: return kz # Complex index of refraction is relative to the incident medium. # We can get the same effect using kz_rel^2 = kz^2 + 4*pi*rho_o # in place of kz^2, and ignoring rho_o kz_sq = kz**2 + 4e-6 * np.pi * rho[:, 0] k = kz # According to Heavens, the initial matrix should be [ 1 F; F 1], # which we do by setting B=I and M0 to [1 F; F 1]. An extra matrix # multiply versus some coding convenience. B11 = 1 B22 = 1 B21 = 0 B12 = 0 for i in range(0, len(depth) - 1): k_next = np.sqrt(kz_sq - 4e-6 * np.pi * (rho[:, i + 1] + 1j * irho[:, i + 1])) F = (k - k_next) / (k + k_next) F *= np.exp(-2 * k * k_next * sigma[i]**2) #print "==== layer",i #print "kz:", kz #print "k:", k #print "k_next:",k_next #print "F:",F #print "rho:",rho[:,i+1] #print "irho:",irho[:,i+1] #print "d:",depth[i],"sigma:",sigma[i] M11 = np.exp(1j * k * depth[i]) if i > 0 else 1 M22 = np.exp(-1j * k * depth[i]) if i > 0 else 1 M21 = F * M11 M12 = F * M22 C1 = B11 * M11 + B21 * M12 C2 = B11 * M21 + B21 * M22 B11 = C1 B21 = C2 C1 = B12 * M11 + B22 * M12 C2 = B12 * M21 + B22 * M22 B12 = C1 B22 = C2 k = k_next r = B12 / B11 return np.absolute(r)**2 if np.isscalar(kz): kz = np.asarray([kz], 'd') m = len(depth) # Make everything into arrays depth = np.asarray(depth, 'd') rho = np.asarray(rho, 'd') irho = irho * np.ones_like(rho) if np.isscalar(irho) else np.asarray( irho, 'd') sigma = sigma * np.ones(m - 1, 'd') if np.isscalar(sigma) else np.asarray( sigma, 'd') # Repeat rho,irho columns as needed if len(rho.shape) == 1: rho = rho[None, :] irho = irho[None, :] return calc(kz, depth, rho, irho, sigma)
[docs]def SmearAbeles(kz, dq, depth, rho, irho=0, sigma=0): y = abeles(kz, depth, rho, irho, sigma) s = dq / 2. y += 0.1354 * (abeles(kz + 2 * s, depth, rho, irho, sigma) + abeles(kz - 2 * s, depth, rho, irho, sigma)) y += 0.24935 * (abeles(kz - 5 * s / 3., depth, rho, irho, sigma) + abeles(kz + 5 * s / 3., depth, rho, irho, sigma)) y += 0.4111 * (abeles(kz - 4 * s / 3., depth, rho, irho, sigma) + abeles(kz + 4 * s / 3., depth, rho, irho, sigma)) y += 0.60653 * (abeles(kz - s, depth, rho, irho, sigma) + abeles(kz + s, depth, rho, irho, sigma)) y += 0.80074 * (abeles(kz - 2 * s / 3., depth, rho, irho, sigma) + abeles(kz - 2 * s / 3., depth, rho, irho, sigma)) y += 0.94596 * (abeles(kz - s / 3., depth, rho, irho, sigma) + abeles(kz - s / 3., depth, rho, irho, sigma)) y *= 0.137023 return y
[docs]def makeRefdFFT(Limits, Profile): G2fil.G2Print('make fft') Q, Io = Profile[:2] Qmin = Limits[1][0] Qmax = Limits[1][1] iBeg = np.searchsorted(Q, Qmin) iFin = np.searchsorted(Q, Qmax) + 1 #include last point Qf = np.linspace(0., Q[iFin - 1], 5000) QI = si.interp1d(Q[iBeg:iFin], Io[iBeg:iFin], bounds_error=False, fill_value=0.0) If = QI(Qf) * Qf**4 R = np.linspace(0., 5000., 5000) F = fft.rfft(If) return R, F
################################################################################ # Stacking fault simulation codes ################################################################################
[docs]def GetStackParms(Layers): Parms = [] #cell parms if Layers['Laue'] in ['-3', '-3m', '4/m', '4/mmm', '6/m', '6/mmm']: Parms.append('cellA') Parms.append('cellC') else: Parms.append('cellA') Parms.append('cellB') Parms.append('cellC') if Layers['Laue'] != 'mmm': Parms.append('cellG') #Transition parms for iY in range(len(Layers['Layers'])): for iX in range(len(Layers['Layers'])): Parms.append('TransP;%d;%d' % (iY, iX)) Parms.append('TransX;%d;%d' % (iY, iX)) Parms.append('TransY;%d;%d' % (iY, iX)) Parms.append('TransZ;%d;%d' % (iY, iX)) return Parms
[docs]def StackSim(Layers, ctrls, scale=0., background={}, limits=[], inst={}, profile=[]): '''Simulate powder or selected area diffraction pattern from stacking faults using DIFFaX :param dict Layers: dict with following items :: {'Laue':'-1','Cell':[False,1.,1.,1.,90.,90.,90,1.], 'Width':[[10.,10.],[False,False]],'Toler':0.01,'AtInfo':{}, 'Layers':[],'Stacking':[],'Transitions':[]} :param str ctrls: controls string to be written on DIFFaX controls.dif file :param float scale: scale factor :param dict background: background parameters :param list limits: min/max 2-theta to be calculated :param dict inst: instrument parameters dictionary :param list profile: powder pattern data Note that parameters all updated in place ''' import atmdata path = sys.path for name in path: if 'bin' in name: DIFFaX = name + '/DIFFaX.exe' G2fil.G2Print(' Execute ' + DIFFaX) break # make form factor file that DIFFaX wants - atom types are GSASII style sf = open('data.sfc', 'w') sf.write('GSASII special form factor file for DIFFaX\n\n') atTypes = list(Layers['AtInfo'].keys()) if 'H' not in atTypes: atTypes.insert(0, 'H') for atType in atTypes: if atType == 'H': blen = -.3741 else: blen = Layers['AtInfo'][atType]['Isotopes']['Nat. Abund.']['SL'][0] Adat = atmdata.XrayFF[atType] text = '%4s' % (atType.ljust(4)) for i in range(4): text += '%11.6f%11.6f' % (Adat['fa'][i], Adat['fb'][i]) text += '%11.6f%11.6f' % (Adat['fc'], blen) text += '%3d\n' % (Adat['Z']) sf.write(text) sf.close() #make DIFFaX control.dif file - future use GUI to set some of these flags cf = open('control.dif', 'w') if ctrls == '0\n0\n3\n' or ctrls == '0\n1\n3\n': x0 = profile[0] iBeg = np.searchsorted(x0, limits[0]) iFin = np.searchsorted(x0, limits[1]) + 1 if iFin - iBeg > 20000: iFin = iBeg + 20000 Dx = (x0[iFin] - x0[iBeg]) / (iFin - iBeg) cf.write('GSASII-DIFFaX.dat\n' + ctrls) cf.write('%.6f %.6f %.6f\n1\n1\nend\n' % (x0[iBeg], x0[iFin], Dx)) else: cf.write('GSASII-DIFFaX.dat\n' + ctrls) inst = { 'Type': [ 'XSC', 'XSC', ] } cf.close() #make DIFFaX data file df = open('GSASII-DIFFaX.dat', 'w') df.write('INSTRUMENTAL\n') if 'X' in inst['Type'][0]: df.write('X-RAY\n') elif 'N' in inst['Type'][0]: df.write('NEUTRON\n') if ctrls == '0\n0\n3\n' or ctrls == '0\n1\n3\n': df.write('%.4f\n' % (G2mth.getMeanWave(inst))) U = ateln2 * inst['U'][1] / 10000. V = ateln2 * inst['V'][1] / 10000. W = ateln2 * inst['W'][1] / 10000. HWHM = U * nptand(x0[iBeg:iFin] / 2.)**2 + V * nptand( x0[iBeg:iFin] / 2.) + W HW = np.sqrt(np.mean(HWHM)) # df.write('PSEUDO-VOIGT 0.015 -0.0036 0.009 0.605 TRIM\n') if 'Mean' in Layers['selInst']: df.write('GAUSSIAN %.6f TRIM\n' % (HW)) #fast option - might not really matter elif 'Gaussian' in Layers['selInst']: df.write('GAUSSIAN %.6f %.6f %.6f TRIM\n' % (U, V, W)) #slow - make a GUI option? else: df.write('None\n') else: df.write('0.10\nNone\n') df.write('STRUCTURAL\n') a, b, c = Layers['Cell'][1:4] gam = Layers['Cell'][6] df.write('%.4f %.4f %.4f %.3f\n' % (a, b, c, gam)) laue = Layers['Laue'] if laue == '2/m(ab)': laue = '2/m(1)' elif laue == '2/m(c)': laue = '2/m(2)' if 'unknown' in Layers['Laue']: df.write('%s %.3f\n' % (laue, Layers['Toler'])) else: df.write('%s\n' % (laue)) df.write('%d\n' % (len(Layers['Layers']))) if Layers['Width'][0][0] < 1. or Layers['Width'][0][1] < 1.: df.write('%.1f %.1f\n' % (Layers['Width'][0][0] * 10000., Layers['Width'][0][0] * 10000.)) #mum to A layerNames = [] for layer in Layers['Layers']: layerNames.append(layer['Name']) for il, layer in enumerate(Layers['Layers']): if layer['SameAs']: df.write('LAYER %d = %d\n' % (il + 1, layerNames.index(layer['SameAs']) + 1)) continue df.write('LAYER %d\n' % (il + 1)) if '-1' in layer['Symm']: df.write('CENTROSYMMETRIC\n') else: df.write('NONE\n') for ia, atom in enumerate(layer['Atoms']): [name, atype, x, y, z, frac, Uiso] = atom if '-1' in layer['Symm'] and [x, y, z] == [0., 0., 0.]: frac /= 2. df.write('%4s %3d %.5f %.5f %.5f %.4f %.2f\n' % (atype.ljust(6), ia, x, y, z, 78.9568 * Uiso, frac)) df.write('STACKING\n') df.write('%s\n' % (Layers['Stacking'][0])) if 'recursive' in Layers['Stacking'][0]: df.write('%s\n' % Layers['Stacking'][1]) else: if 'list' in Layers['Stacking'][1]: Slen = len(Layers['Stacking'][2]) iB = 0 iF = 0 while True: iF += 68 if iF >= Slen: break iF = min(iF, Slen) df.write('%s\n' % (Layers['Stacking'][2][iB:iF])) iB = iF else: df.write('%s\n' % Layers['Stacking'][1]) df.write('TRANSITIONS\n') for iY in range(len(Layers['Layers'])): sumPx = 0. for iX in range(len(Layers['Layers'])): p, dx, dy, dz = Layers['Transitions'][iY][iX][:4] p = round(p, 3) df.write('%.3f %.5f %.5f %.5f\n' % (p, dx, dy, dz)) sumPx += p if sumPx != 1.0: #this has to be picky since DIFFaX is. G2fil.G2Print( 'ERROR - Layer probabilities sum to %.3f DIFFaX will insist it = 1.0' % sumPx) df.close() os.remove('data.sfc') os.remove('control.dif') os.remove('GSASII-DIFFaX.dat') return df.close() time0 = time.time() try: subp.call(DIFFaX) except OSError: G2fil.G2Print('DIFFax.exe is not available for this platform', mode='warn') G2fil.G2Print(' DIFFaX time = %.2fs' % (time.time() - time0)) if os.path.exists('GSASII-DIFFaX.spc'): Xpat = np.loadtxt('GSASII-DIFFaX.spc').T iFin = iBeg + Xpat.shape[1] bakType, backDict, backVary = SetBackgroundParms(background) backDict['Lam1'] = G2mth.getWave(inst) profile[4][iBeg:iFin] = getBackground('', backDict, bakType, inst['Type'][0], profile[0][iBeg:iFin])[0] profile[3][iBeg:iFin] = Xpat[-1] * scale + profile[4][iBeg:iFin] if not np.any(profile[1]): #fill dummy data x,y,w,yc,yb,yd rv = st.poisson(profile[3][iBeg:iFin]) profile[1][iBeg:iFin] = rv.rvs() Z = np.ones_like(profile[3][iBeg:iFin]) Z[1::2] *= -1 profile[1][iBeg:iFin] = profile[3][iBeg:iFin] + np.abs( profile[1][iBeg:iFin] - profile[3][iBeg:iFin]) * Z profile[2][iBeg:iFin] = np.where(profile[1][iBeg:iFin] > 0., 1. / profile[1][iBeg:iFin], 1.0) profile[5][iBeg:iFin] = profile[1][iBeg:iFin] - profile[3][iBeg:iFin] #cleanup files.. os.remove('GSASII-DIFFaX.spc') elif os.path.exists('GSASII-DIFFaX.sadp'): Sadp = np.fromfile('GSASII-DIFFaX.sadp', '>u2') Sadp = np.reshape(Sadp, (256, -1)) Layers['Sadp']['Img'] = Sadp os.remove('GSASII-DIFFaX.sadp') os.remove('data.sfc') os.remove('control.dif') os.remove('GSASII-DIFFaX.dat')
[docs]def SetPWDRscan(inst, limits, profile): wave = G2mth.getMeanWave(inst) x0 = profile[0] iBeg = np.searchsorted(x0, limits[0]) iFin = np.searchsorted(x0, limits[1]) if iFin - iBeg > 20000: iFin = iBeg + 20000 Dx = (x0[iFin] - x0[iBeg]) / (iFin - iBeg) pyx.pygetinst(wave, x0[iBeg], x0[iFin], Dx) return iFin - iBeg
[docs]def SetStackingSF(Layers, debug): # Load scattering factors into DIFFaX arrays import atmdata atTypes = Layers['AtInfo'].keys() aTypes = [] for atype in atTypes: aTypes.append('%4s' % (atype.ljust(4))) SFdat = [] for atType in atTypes: Adat = atmdata.XrayFF[atType] SF = np.zeros(9) SF[:8:2] = Adat['fa'] SF[1:8:2] = Adat['fb'] SF[8] = Adat['fc'] SFdat.append(SF) SFdat = np.array(SFdat) pyx.pyloadscf(len(atTypes), aTypes, SFdat.T, debug)
[docs]def SetStackingClay(Layers, Type): # Controls #rand.seed() ranSeed = rand.randint(1, 2**16 - 1) try: laueId = [ '-1', '2/m(ab)', '2/m(c)', 'mmm', '-3', '-3m', '4/m', '4/mmm', '6/m', '6/mmm' ].index(Layers['Laue']) + 1 except ValueError: #for 'unknown' laueId = -1 if 'SADP' in Type: planeId = ['h0l', '0kl', 'hhl', 'h-hl'].index( Layers['Sadp']['Plane']) + 1 lmax = int(Layers['Sadp']['Lmax']) else: planeId = 0 lmax = 0 # Sequences StkType = ['recursive', 'explicit'].index(Layers['Stacking'][0]) try: StkParm = ['infinite', 'random', 'list'].index(Layers['Stacking'][1]) except ValueError: StkParm = -1 if StkParm == 2: #list StkSeq = [int(val) for val in Layers['Stacking'][2].split()] Nstk = len(StkSeq) else: Nstk = 1 StkSeq = [ 0, ] if StkParm == -1: StkParm = int(Layers['Stacking'][1]) Wdth = Layers['Width'][0] mult = 1 controls = [laueId, planeId, lmax, mult, StkType, StkParm, ranSeed] LaueSym = Layers['Laue'].ljust(12) pyx.pygetclay(controls, LaueSym, Wdth, Nstk, StkSeq) return laueId, controls
[docs]def SetCellAtoms(Layers): Cell = Layers['Cell'][1:4] + Layers['Cell'][6:7] # atoms in layers atTypes = list(Layers['AtInfo'].keys()) AtomXOU = [] AtomTp = [] LayerSymm = [] LayerNum = [] layerNames = [] Natm = 0 Nuniq = 0 for layer in Layers['Layers']: layerNames.append(layer['Name']) for il, layer in enumerate(Layers['Layers']): if layer['SameAs']: LayerNum.append(layerNames.index(layer['SameAs']) + 1) continue else: LayerNum.append(il + 1) Nuniq += 1 if '-1' in layer['Symm']: LayerSymm.append(1) else: LayerSymm.append(0) for ia, atom in enumerate(layer['Atoms']): [name, atype, x, y, z, frac, Uiso] = atom Natm += 1 AtomTp.append('%4s' % (atype.ljust(4))) Ta = atTypes.index(atype) + 1 AtomXOU.append([ float(Nuniq), float(ia + 1), float(Ta), x, y, z, frac, Uiso * 78.9568 ]) AtomXOU = np.array(AtomXOU) Nlayers = len(layerNames) pyx.pycellayer(Cell, Natm, AtomTp, AtomXOU.T, Nuniq, LayerSymm, Nlayers, LayerNum) return Nlayers
[docs]def SetStackingTrans(Layers, Nlayers): # Transitions TransX = [] TransP = [] for Ytrans in Layers['Transitions']: TransP.append([trans[0] for trans in Ytrans]) #get just the numbers TransX.append([trans[1:4] for trans in Ytrans]) #get just the numbers TransP = np.array(TransP, dtype='float').T TransX = np.array(TransX, dtype='float') # GSASIIpath.IPyBreak() pyx.pygettrans(Nlayers, TransP, TransX)
[docs]def CalcStackingPWDR(Layers, scale, background, limits, inst, profile, debug): # Scattering factors SetStackingSF(Layers, debug) # Controls & sequences laueId, controls = SetStackingClay(Layers, 'PWDR') # cell & atoms Nlayers = SetCellAtoms(Layers) Volume = Layers['Cell'][7] # Transitions SetStackingTrans(Layers, Nlayers) # PWDR scan Nsteps = SetPWDRscan(inst, limits, profile) # result as Spec x0 = profile[0] profile[3] = np.zeros(len(profile[0])) profile[4] = np.zeros(len(profile[0])) profile[5] = np.zeros(len(profile[0])) iBeg = np.searchsorted(x0, limits[0]) iFin = np.searchsorted(x0, limits[1]) + 1 if iFin - iBeg > 20000: iFin = iBeg + 20000 Nspec = 20001 spec = np.zeros(Nspec, dtype='double') time0 = time.time() pyx.pygetspc(controls, Nspec, spec) G2fil.G2Print(' GETSPC time = %.2fs' % (time.time() - time0)) time0 = time.time() U = ateln2 * inst['U'][1] / 10000. V = ateln2 * inst['V'][1] / 10000. W = ateln2 * inst['W'][1] / 10000. HWHM = U * nptand(x0[iBeg:iFin] / 2.)**2 + V * nptand( x0[iBeg:iFin] / 2.) + W HW = np.sqrt(np.mean(HWHM)) BrdSpec = np.zeros(Nsteps) if 'Mean' in Layers['selInst']: pyx.pyprofile(U, V, W, HW, 1, Nsteps, BrdSpec) elif 'Gaussian' in Layers['selInst']: pyx.pyprofile(U, V, W, HW, 4, Nsteps, BrdSpec) else: BrdSpec = spec[:Nsteps] BrdSpec /= Volume iFin = iBeg + Nsteps bakType, backDict, backVary = SetBackgroundParms(background) backDict['Lam1'] = G2mth.getWave(inst) profile[4][iBeg:iFin] = getBackground('', backDict, bakType, inst['Type'][0], profile[0][iBeg:iFin])[0] profile[3][iBeg:iFin] = BrdSpec * scale + profile[4][iBeg:iFin] if not np.any(profile[1]): #fill dummy data x,y,w,yc,yb,yd try: rv = st.poisson(profile[3][iBeg:iFin]) profile[1][iBeg:iFin] = rv.rvs() except ValueError: profile[1][iBeg:iFin] = profile[3][iBeg:iFin] Z = np.ones_like(profile[3][iBeg:iFin]) Z[1::2] *= -1 profile[1][iBeg:iFin] = profile[3][iBeg:iFin] + np.abs( profile[1][iBeg:iFin] - profile[3][iBeg:iFin]) * Z profile[2][iBeg:iFin] = np.where(profile[1][iBeg:iFin] > 0., 1. / profile[1][iBeg:iFin], 1.0) profile[5][iBeg:iFin] = profile[1][iBeg:iFin] - profile[3][iBeg:iFin] G2fil.G2Print(' Broadening time = %.2fs' % (time.time() - time0))
[docs]def CalcStackingSADP(Layers, debug): # Scattering factors SetStackingSF(Layers, debug) # Controls & sequences laueId, controls = SetStackingClay(Layers, 'SADP') # cell & atoms Nlayers = SetCellAtoms(Layers) # Transitions SetStackingTrans(Layers, Nlayers) # result as Sadp Nspec = 20001 spec = np.zeros(Nspec, dtype='double') time0 = time.time() hkLim, Incr, Nblk = pyx.pygetsadp(controls, Nspec, spec) Sapd = np.zeros((256, 256)) iB = 0 for i in range(hkLim): iF = iB + Nblk p1 = 127 + int(i * Incr) p2 = 128 - int(i * Incr) if Nblk == 128: if i: Sapd[128:, p1] = spec[iB:iF] Sapd[:128, p1] = spec[iF:iB:-1] Sapd[128:, p2] = spec[iB:iF] Sapd[:128, p2] = spec[iF:iB:-1] else: if i: Sapd[:, p1] = spec[iB:iF] Sapd[:, p2] = spec[iB:iF] iB += Nblk Layers['Sadp']['Img'] = Sapd G2fil.G2Print(' GETSAD time = %.2fs' % (time.time() - time0))
############################################################################### #### Maximum Entropy Method - Dysnomia ###############################################################################
[docs]def makePRFfile(data, MEMtype): ''' makes Dysnomia .prf control file from Dysnomia GUI controls :param dict data: GSAS-II phase data :param int MEMtype: 1 for neutron data with negative scattering lengths 0 otherwise :returns str: name of Dysnomia control file ''' generalData = data['General'] pName = generalData['Name'].replace(' ', '_') DysData = data['Dysnomia'] prfName = pName + '.prf' prf = open(prfName, 'w') prf.write('$PREFERENCES\n') prf.write(pName + '.mem\n') #or .fos? prf.write(pName + '.out\n') prf.write(pName + '.pgrid\n') prf.write(pName + '.fba\n') prf.write(pName + '_eps.raw\n') prf.write('%d\n' % MEMtype) if DysData['DenStart'] == 'uniform': prf.write('0\n') else: prf.write('1\n') if DysData['Optimize'] == 'ZSPA': prf.write('0\n') else: prf.write('1\n') prf.write('1\n') if DysData['Lagrange'][0] == 'user': prf.write('0\n') else: prf.write('1\n') prf.write('%.4f %d\n' % (DysData['Lagrange'][1], DysData['wt pwr'])) prf.write('%.3f\n' % DysData['Lagrange'][2]) prf.write('%.2f\n' % DysData['E_factor']) prf.write('1\n') prf.write('0\n') prf.write('%d\n' % DysData['Ncyc']) prf.write('1\n') prf.write('1 0 0 0 0 0 0 0\n') if DysData['prior'] == 'uniform': prf.write('0\n') else: prf.write('1\n') prf.close() return prfName
[docs]def makeMEMfile(data, reflData, MEMtype, DYSNOMIA): ''' make Dysnomia .mem file of reflection data, etc. :param dict data: GSAS-II phase data :param list reflData: GSAS-II reflection data :param int MEMtype: 1 for neutron data with negative scattering lengths 0 otherwise :param str DYSNOMIA: path to dysnomia.exe ''' DysData = data['Dysnomia'] generalData = data['General'] cell = generalData['Cell'][1:7] A = G2lat.cell2A(cell) SGData = generalData['SGData'] pName = generalData['Name'].replace(' ', '_') memName = pName + '.mem' Map = generalData['Map'] Type = Map['Type'] UseList = Map['RefList'] mem = open(memName, 'w') mem.write('%s\n' % (generalData['Name'] + ' from ' + UseList[0])) a, b, c, alp, bet, gam = cell mem.write('%10.5f%10.5f%10.5f%10.5f%10.5f%10.5f\n' % (a, b, c, alp, bet, gam)) mem.write(' 0.0000000 0.0000000 -1 0 0 0 P\n' ) #dummy PO stuff SGSym = generalData['SGData']['SpGrp'] try: SGId = G2spc.spgbyNum.index(SGSym) except ValueError: return False org = 1 if SGSym in G2spc.spg2origins: org = 2 mapsize = Map['rho'].shape sumZ = 0. sumpos = 0. sumneg = 0. mem.write('%5d%5d%5d%5d%5d\n' % (SGId, org, mapsize[0], mapsize[1], mapsize[2])) for atm in generalData['NoAtoms']: Nat = generalData['NoAtoms'][atm] AtInfo = G2elem.GetAtomInfo(atm) sumZ += Nat * AtInfo['Z'] isotope = generalData['Isotope'][atm] blen = generalData['Isotopes'][atm][isotope]['SL'][0] if blen < 0.: sumneg += blen * Nat else: sumpos += blen * Nat if 'X' in Type: mem.write('%10.2f 0.001\n' % sumZ) elif 'N' in Type and MEMtype: mem.write('%10.3f%10.3f 0.001\n' % (sumpos, sumneg)) else: mem.write('%10.3f 0.001\n' % sumpos) dmin = DysData['MEMdmin'] TOFlam = 2.0 * dmin * npsind(80.0) refSet = G2lat.GenHLaue(dmin, SGData, A) #list of h,k,l,d refDict = {'%d %d %d' % (ref[0], ref[1], ref[2]): ref for ref in refSet} refs = [] prevpos = 0. for ref in reflData: if ref[3] < 0: continue if 'T' in Type: h, k, l, mult, dsp, pos, sig, gam, Fobs, Fcalc, phase, x, x, x, x, prfo = ref[: 16] s = np.sqrt(max(sig, 0.0001)) #var -> sig in deg FWHM = getgamFW(gam, s) if dsp < dmin: continue theta = npasind(TOFlam / (2. * dsp)) FWHM *= nptand(theta) / pos pos = 2. * theta else: h, k, l, mult, dsp, pos, sig, gam, Fobs, Fcalc, phase, x, prfo = ref[: 13] g = gam / 100. #centideg -> deg s = np.sqrt(max(sig, 0.0001)) / 100. #var -> sig in deg FWHM = getgamFW(g, s) delt = pos - prevpos refs.append([h, k, l, mult, pos, FWHM, Fobs, phase, delt]) prevpos = pos ovlp = DysData['overlap'] refs1 = [] refs2 = [] nref2 = 0 iref = 0 Nref = len(refs) start = False while iref < Nref - 1: if refs[iref + 1][-1] < ovlp * refs[iref][5]: if refs[iref][-1] > ovlp * refs[iref][5]: refs2.append([]) start = True if nref2 == len(refs2): refs2.append([]) refs2[nref2].append(refs[iref]) else: if start: refs2[nref2].append(refs[iref]) start = False nref2 += 1 else: refs1.append(refs[iref]) iref += 1 if start: refs2[nref2].append(refs[iref]) else: refs1.append(refs[iref]) mem.write('%5d\n' % len(refs1)) for ref in refs1: h, k, l = ref[:3] hkl = '%d %d %d' % (h, k, l) if hkl in refDict: del refDict[hkl] Fobs = np.sqrt(ref[6]) mem.write('%5d%5d%5d%10.3f%10.3f%10.3f\n' % (h, k, l, Fobs * npcosd(ref[7]), Fobs * npsind(ref[7]), max(0.01 * Fobs, 0.1))) while True and nref2: if not len(refs2[-1]): del refs2[-1] else: break mem.write('%5d\n' % len(refs2)) for iref2, ref2 in enumerate(refs2): mem.write('#%5d\n' % iref2) mem.write('%5d\n' % len(ref2)) Gsum = 0. Msum = 0 for ref in ref2: Gsum += ref[6] * ref[3] Msum += ref[3] G = np.sqrt(Gsum / Msum) h, k, l = ref2[0][:3] hkl = '%d %d %d' % (h, k, l) if hkl in refDict: del refDict[hkl] mem.write('%5d%5d%5d%10.3f%10.3f%5d\n' % (h, k, l, G, max(0.01 * G, 0.1), ref2[0][3])) for ref in ref2[1:]: h, k, l, m = ref[:4] mem.write('%5d%5d%5d%5d\n' % (h, k, l, m)) hkl = '%d %d %d' % (h, k, l) if hkl in refDict: del refDict[hkl] if len(refDict): mem.write('%d\n' % len(refDict)) for hkl in list(refDict.keys()): h, k, l = refDict[hkl][:3] mem.write('%5d%5d%5d\n' % (h, k, l)) else: mem.write('0\n') mem.close() return True
[docs]def MEMupdateReflData(prfName, data, reflData): ''' Update reflection data with new Fosq, phase result from Dysnomia :param str prfName: phase.mem file name :param list reflData: GSAS-II reflection data ''' generalData = data['General'] Map = generalData['Map'] Type = Map['Type'] cell = generalData['Cell'][1:7] A = G2lat.cell2A(cell) reflDict = {} newRefs = [] for iref, ref in enumerate(reflData): if ref[3] > 0: newRefs.append(ref) reflDict[hash('%5d%5d%5d' % (ref[0], ref[1], ref[2]))] = iref fbaName = os.path.splitext(prfName)[0] + '.fba' try: fba = open(fbaName, 'r') except FileNotFoundError: return False fba.readline() Nref = int(fba.readline()[:-1]) fbalines = fba.readlines() for line in fbalines[:Nref]: info = line.split() h = int(info[0]) k = int(info[1]) l = int(info[2]) FoR = float(info[3]) FoI = float(info[4]) Fosq = FoR**2 + FoI**2 phase = npatan2d(FoI, FoR) try: refId = reflDict[hash('%5d%5d%5d' % (h, k, l))] except KeyError: #added reflections at end skipped d = float(1 / np.sqrt(G2lat.calc_rDsq([h, k, l], A))) if 'T' in Type: newRefs.append([ h, k, l, -1, d, 0., 0.01, 1.0, Fosq, Fosq, phase, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0 ]) else: newRefs.append([ h, k, l, -1, d, 0., 0.01, 1.0, Fosq, Fosq, phase, 1.0, 1.0, 1.0, 1.0 ]) continue newRefs[refId][8] = Fosq newRefs[refId][10] = phase newRefs = np.array(newRefs) return True, newRefs
#testing data NeedTestData = True
[docs]def TestData(): 'needs a doc string' # global NeedTestData global bakType bakType = 'chebyschev' global xdata xdata = np.linspace(4.0, 40.0, 36000) global parmDict0 parmDict0 = { 'pos0': 5.6964, 'int0': 8835.8, 'sig0': 1.0, 'gam0': 1.0, 'pos1': 11.4074, 'int1': 3922.3, 'sig1': 1.0, 'gam1': 1.0, 'pos2': 20.6426, 'int2': 1573.7, 'sig2': 1.0, 'gam2': 1.0, 'pos3': 26.9568, 'int3': 925.1, 'sig3': 1.0, 'gam3': 1.0, 'U': 1.163, 'V': -0.605, 'W': 0.093, 'X': 0.0, 'Y': 2.183, 'Z': 0.0, 'SH/L': 0.002, 'Back0': 5.384, 'Back1': -0.015, 'Back2': .004, } global parmDict1 parmDict1 = { 'pos0': 13.4924, 'int0': 48697.6, 'sig0': 1.0, 'gam0': 1.0, 'pos1': 23.4360, 'int1': 43685.5, 'sig1': 1.0, 'gam1': 1.0, 'pos2': 27.1152, 'int2': 123712.6, 'sig2': 1.0, 'gam2': 1.0, 'pos3': 33.7196, 'int3': 65349.4, 'sig3': 1.0, 'gam3': 1.0, 'pos4': 36.1119, 'int4': 115829.8, 'sig4': 1.0, 'gam4': 1.0, 'pos5': 39.0122, 'int5': 6916.9, 'sig5': 1.0, 'gam5': 1.0, 'U': 22.75, 'V': -17.596, 'W': 10.594, 'X': 1.577, 'Y': 5.778, 'Z': 0.0, 'SH/L': 0.002, 'Back0': 36.897, 'Back1': -0.508, 'Back2': .006, 'Lam1': 1.540500, 'Lam2': 1.544300, 'I(L2)/I(L1)': 0.5, } global parmDict2 parmDict2 = { 'pos0': 5.7, 'int0': 1000.0, 'sig0': 0.5, 'gam0': 0.5, 'U': 2., 'V': -2., 'W': 5., 'X': 0.5, 'Y': 0.5, 'Z': 0.0, 'SH/L': 0.02, 'Back0': 5., 'Back1': -0.02, 'Back2': .004, # 'Lam1':1.540500,'Lam2':1.544300,'I(L2)/I(L1)':0.5, } global varyList varyList = []
[docs]def test0(): if NeedTestData: TestData() gplot = plotter.add('FCJ-Voigt, 11BM').gca() gplot.plot(xdata, getBackground('', parmDict0, bakType, 'PXC', xdata)[0]) gplot.plot(xdata, getPeakProfile(parmDict0, xdata, varyList, bakType)) fplot = plotter.add('FCJ-Voigt, Ka1+2').gca() fplot.plot(xdata, getBackground('', parmDict1, bakType, 'PXC', xdata)[0]) fplot.plot(xdata, getPeakProfile(parmDict1, xdata, varyList, bakType))
[docs]def test1(): if NeedTestData: TestData() time0 = time.time() for i in range(100): getPeakProfile(parmDict1, xdata, varyList, bakType) G2fil.G2Print('100+6*Ka1-2 peaks=1200 peaks %.2f' % time.time() - time0)
[docs]def test2(name, delt): if NeedTestData: TestData() varyList = [ name, ] xdata = np.linspace(5.6, 5.8, 400) hplot = plotter.add('derivatives test for ' + name).gca() hplot.plot(xdata, getPeakProfileDerv(parmDict2, xdata, varyList, bakType)[0]) y0 = getPeakProfile(parmDict2, xdata, varyList, bakType) parmDict2[name] += delt y1 = getPeakProfile(parmDict2, xdata, varyList, bakType) hplot.plot(xdata, (y1 - y0) / delt, 'r+')
[docs]def test3(name, delt): if NeedTestData: TestData() names = ['pos', 'sig', 'gam', 'shl'] idx = names.index(name) myDict = { 'pos': parmDict2['pos0'], 'sig': parmDict2['sig0'], 'gam': parmDict2['gam0'], 'shl': parmDict2['SH/L'] } xdata = np.linspace(5.6, 5.8, 800) dx = xdata[1] - xdata[0] hplot = plotter.add('derivatives test for ' + name).gca() hplot.plot( xdata, 100. * dx * getdFCJVoigt3(myDict['pos'], myDict['sig'], myDict['gam'], myDict['shl'], xdata)[idx + 1]) y0 = getFCJVoigt3(myDict['pos'], myDict['sig'], myDict['gam'], myDict['shl'], xdata) myDict[name] += delt y1 = getFCJVoigt3(myDict['pos'], myDict['sig'], myDict['gam'], myDict['shl'], xdata) hplot.plot(xdata, (y1 - y0) / delt, 'r+')
if __name__ == '__main__': import GSASIItestplot as plot global plotter plotter = plot.PlotNotebook() # test0() # for name in ['int0','pos0','sig0','gam0','U','V','W','X','Y','Z','SH/L','I(L2)/I(L1)']: for name, shft in [['int0', 0.1], ['pos0', 0.0001], ['sig0', 0.01], ['gam0', 0.00001], ['U', 0.1], ['V', 0.01], ['W', 0.01], ['X', 0.0001], ['Y', 0.0001], ['Z', 0.0001], ['SH/L', 0.00005]]: test2(name, shft) for name, shft in [['pos', 0.0001], ['sig', 0.01], ['gam', 0.0001], ['shl', 0.00005]]: test3(name, shft) G2fil.G2Print("OK") plotter.StartEventLoop()