farhi
commented
6 years ago Message from David Voneshen - UKRI STFC david.voneshen@stfc.ac.uk on Aug 24th 2018 2:48PM:
Hi Emmanuel,
I am trying to use ResLibCal (latest) and spinw (R1321) within MATLAB 2015b to model some data we recently collected on IN8. My issues seem to be almost identical to the iFit bug #160 reported by Henrik Jacobsen earlier this year.
Unlike Henrik I am able to use ResLibCal and the simple example for sqw_sine3d with some minor modifications. My sine3d script looks like
s=sqw_sine3d(1.); % create a 4D S(q,w) for a cubic pure material, using default parameters
t=ResLibCal(s); % convolute it with a TAS resolution, and open ResLibCal.
w=linspace(0.5,2.0,50); qh=1.1*ones(size(w)); qk=0*qh; ql=0*qh; % an E-scan around the [1 0 0] Bragg
signal0=iData(s, [], qh,qk,ql,w); % the ideal model evaluated along the HKLE line
signal1=iData(t, [], qh,qk,ql,w); % the convoluted model evaluated along the HKLE line
plot(w,signal0.signal,w,signal1.signal*120)When trying to use spinW however I seem to get the exact same issue as Henrik. The minimum code required to reproduce my problem is
s=sqw_spinw()
t=ResLibCal(s);
w=linspace(0,10,101);
qh=1.15*ones(size(w)); qk=0*qh; ql=0*qh;
cut=iData(t, t.p, qh,qk,ql,w)
plot(w,cut.signal)Which produces the following output
1.10 iFit/iData (Aug. 22, 2017) by E.Farhi, P. Willendrup and Y.Debab, (c) ILL/DS/CS <farhi@ill.eu> EUPL. Aug. 22, 2017
Opening iData documentation from
web C:\Users\tyo59427\Documents\projects\spin-cal\Magnetite\iFit\ifit-1.10\Objects\@iData\..\..\\Docs\Models.html#mozTocId908192
Calculation is finished in 00:00:00 (hh:mm:ss).
sqw_spinw: Model atom_1 built using SpinW.
Ground state energy: -4 [mev/spin]
* S. Toth and B. Lake, J. Phys.: Condens. Matter 27, 166002 (2015).
s = iFunc (methods,doc,plot,more...) 4D model:
[Tag] [Dim] [Model] [Parameters 'p']
iF488838 4 Sqw_spinw spin-wav. A HKL spi. ax = ... Gamma Temperature Amplitude J1
ResLibCal_tas_conv4d: Using reference frame: rlu and method: unknown
[R] rlu lattice frame, xyz=[a* b* c*]
conv(Sqw_spinw spin-wave dispersion(HKL) [sqw_spinw], ResLibCal('struct('Title','ResL...')) [ResLibCal_tas_conv4d]
Calculation is finished in 00:00:00 (hh:mm:ss).
Error: Could not evaluate Expression in model Sqw_spinw spin-wave dispersion(HKL) [sqw_spinw] iF488838
this = iFunc (methods,doc,plot,code) 4D model: "Sqw_spinw spin-wave dispersion(HKL) [s..."
Expression: ax = {x y z t};check_vector = cellfun(@(c)(~isempty(c) && length(c)...
Description: A HKL spin-wave dispersion from SpinW package with Gaussian energy width: atom_1 J1
Tag: 'iF488838'
Date: '24-Aug-2018 11:38:50'
Name: 'Sqw_spinw spin-wave dispersion(HKL) [sqw_spinw]'
Parameters: {'Gamma energy broadening around spin-wave modes [meV]' 'Temperature [K]' 'Amplitude' 'J1'}
Guess: [0.3000 0 1 2]
Constraint: [1x1 struct]
Dimension: 4
ParameterValues: [4x1 double]
UserData: [1x1 struct]
Duration: 0
Parameters (4):
p( 1)= Gamma=0.3 % entered as: energy broadening around spin-wave modes [meV]
p( 2)= Temperature=0 % entered as: [K]
p( 3)= Amplitude=1
p( 4)= J1=2
''
'ax = {x y z t};'
'check_vector = cellfun(@(c)(~isempty(c) && length(c) == numel(c)), ax);'
'check_numel = cellfun(@numel, ax);'
'check_orient = zeros(1,4);'
'index = find(check_vector);'
'check_orient(index) = cellfun(@(c)find(size(c)==numel(c)), ax(index));'
'HKL = []; is_event = false; resize_me = [];'
'if all(check_vector) && all(check_numel == check_numel(1)) && all(check_orient == check_orient(1));'
'is_event = true;'
...
'spec = spinwave(this.UserData.spinw, HKL');'
'spec = sw_neutron(spec);'
'this.UserData.maxFreq=max(spec.omega(:));'
'spec = sw_egrid(spec,'component',this.UserData.component,'Evect',t(:)', 'T', p(2));'
'spec = sw_instrument(spec,'dE',p(1),'ki',this.UserData.ki);'
'signal = reshape(spec.swConv,resize_me([4 1:3])); signal=p(3)*permute(signal,[2:4 1]);'
Index exceeds matrix dimensions.
Warning: Control Character '\U' is not valid. See 'doc sprintf' for control characters valid in the format string.
> In iFunc/feval>iFunc_feval_expr (line 428)
In iFunc/feval (line 343)
In iFunc/feval>iFunc_feval_expr (line 413)
In iFunc/feval (line 343)
In iData>iData_iFunc2iData (line 247)
In iData (line 173)
Warning: iData/setaxis: the Alias Axis_1 used to define axis rank 1 does not exist in object out iD592717 "".
Defining it.
cut = iData (methods,doc,plot,plot(log),values,more...) 4D object:
[Tag] [Dimension] [Title] [Last command] [Label/DisplayName]
iD592717 [1 101] 'conv(Sqw_spinw spin-wave di... "Data ..."' iD592717=setalias(iD... conv(Sqw_s/conv(Sqw_s...I can if you wish also send my actual spinw model and code but the error is the same. I’ve also tried using the internal version of spinw but it produces the same error. Is there a fix for this? Or will I need to move to use a different package?
Best wishes, David Voneshen P.S. the requested smile J
Bug found by H. Jacobsen Feb 27th 2018:
which produces the output:
Also, it looks like the sign of the coupling parameters J gets positive disregarding how they are input (-J). This could be the issue.