Skip to content

Commit

Permalink
Refs #6670 code, test, and wiki
Browse files Browse the repository at this point in the history
  • Loading branch information
@jmborr
jmborr committed May 23, 2013
1 parent 46365f5 commit 048111b
Show file tree
Hide file tree
Showing 4 changed files with 264 additions and 41 deletions.
3 changes: 3 additions & 0 deletions Code/Mantid/Framework/Algorithms/CMakeLists.txt
View file
Expand Up @@ -130,6 +130,7 @@ set ( SRC_FILES
src/MergeRuns.cpp
src/Min.cpp
src/Minus.cpp
src/ModeratorTzero.cpp
src/ModeratorTzeroLinear.cpp
src/MonteCarloAbsorption.cpp
src/MultipleScatteringCylinderAbsorption.cpp
Expand Down Expand Up @@ -344,6 +345,7 @@ set ( INC_FILES
inc/MantidAlgorithms/MergeRuns.h
inc/MantidAlgorithms/Min.h
inc/MantidAlgorithms/Minus.h
inc/MantidAlgorithms/ModeratorTzero.h
inc/MantidAlgorithms/ModeratorTzeroLinear.h
inc/MantidAlgorithms/MonteCarloAbsorption.h
inc/MantidAlgorithms/MultipleScatteringCylinderAbsorption.h
Expand Down Expand Up @@ -554,6 +556,7 @@ set ( TEST_FILES
MedianDetectorTestTest.h
MergeRunsTest.h
MinusTest.h
ModeratorTzeroTest.h
ModeratorTzeroLinearTest.h
MonteCarloAbsorptionTest.h
MultipleScatteringCylinderAbsorptionTest.h
Expand Down
2 changes: 1 addition & 1 deletion Code/Mantid/Framework/Algorithms/inc/MantidAlgorithms/ModeratorTzero.h
View file
Expand Up @@ -59,7 +59,7 @@ namespace Algorithms
File change history is stored at: <https://github.com/mantidproject/mantid>
Code Documentation is available at: <http://doxygen.mantidproject.org>
*/
class DLLExport ModeratorTzero : public Mantid::API::Algorithm
class DLLExport ModeratorTzero: public Mantid::API::Algorithm
{
public:
/// (Empty) Constructor
Expand Down
93 changes: 53 additions & 40 deletions Code/Mantid/Framework/Algorithms/src/ModeratorTzero.cpp
View file
@@ -1,27 +1,35 @@
/*WIKI*
Corrects the time of flight (TOF) by a time offset that is dependent on the energy of the neutron after passing through the moderator.
A heuristic formula for the correction is stored in the instrument definition file. Below is shown the entry in the instrument file for the VISION beamline:
<!-- formula for t0 calculation. See http://muparser.sourceforge.net/mup_features.html#idDef2 for available operators-->
<parameter name="t0_formula" type="string">
<value val="34.746 - 0.166672*incidentEnergy + 0.00020538*incidentEnergy^(2.0)" />
</parameter>
The recorded TOF = t_0 + t_i + t_f with
t_0: emission time from the moderator
t_1: time from moderator to sample or monitor
t_2: time from sample to detector
This algorithm will replace TOF with TOF' = TOF-t_0 = t_i+t_f
For a direct geometry instrument, the incident energy E_1 is the same for all neutrons. Hence, the moderator emission time is the same for all neutrons.
For an indirect geometry instrument, E_1 is different for each neutron and is not known. However, the final energy E_2 selected by the analyzers is known.
t_0 = func(E_1) , a function of the incident energy
t_1 = L_1/v_1 with L_1 the distance from moderator to sample, and v_1 the initial unknown velocity ( E_1=1/2*m*v_1^2)
t_2 = L_2/v_2 with L_2 the distance from sample to detector, and v_2 is the final fixed velocity ( E_2=1/2*m*v_2^2)
We obtain TOF' in an iterative process, taking into account the fact that the correction t_0 is much smaller than t_i+t_f. Thus
TOF-t_0^(n) = L_1/v_1^(n) + L_2/v_2 , n=0, 1, 2,..
Set t_0^(0)=0 and obtain v_1^(0) from the previous formula. From v_1^(0) we obtain E_1^(0)
Set t_0^(1)=func( E_1^(0) ) and repeat the steps until |t_0^(n+1) - t_0^(n+1)| < 1 microsec. Typically, three to four iterations are needed for convergence.
Corrects the time of flight (TOF) by a time offset that is dependent on the energy of the neutron after passing through the moderator.
A heuristic formula for the correction is stored in the instrument definition file. Below is shown the entry in the instrument file for the VISION beamline:<br />
<nowiki><!-- formula for t0 calculation. See http://muparser.sourceforge.net/mup_features.html#idDef2 for available operators--></nowiki>
<nowiki> <parameter name="t0_formula" type="string"></nowiki>
<nowiki> <value val="34.746 - 0.166672*incidentEnergy + 0.00020538*incidentEnergy^(2.0)" /></nowiki>
<nowiki></parameter></nowiki>
The recorded <math>TOF = t_0 + t_i + t_f</math> with<br />
<math>t_0</math>: emission time from the moderator<br />
<math>t_1</math>: time from moderator to sample or monitor<br />
<math>t_2</math>: time from sample to detector<br />
This algorithm will replace <math>TOF</math> with <math>TOF^* = TOF-t_0 = t_i+t_f</math><br />
For a direct geometry instrument, the incident energy <math>E_1</math> is the same for all neutrons. Hence, the moderator emission time is the same for all neutrons.
For an indirect geometry instrument, <math>E_1</math> is different for each neutron and is not known. However, the final energy <math>E_2</math> selected by the analyzers is known.<br />
<math>t_0 = func(E_1)</math> , a function of the incident energy<br />
<math>t_1 = L_1/v_1</math> with <math>L_1</math> the distance from moderator to sample, and <math>v_1</math> the initial unknown velocity ( <math>E_1=1/2*m*v_1^2</math>)<br />
<math>t_2 = L_2/v_2</math> with <math>L_2</math> the distance from sample to detector, and <math>v_2</math> is the final fixed velocity ( <math>E_2=1/2*m*v_2^2</math>)<br />
We obtain <math>TOF^*</math> as an iterative process, taking into account the fact that the correction <math>t_0</math> is much smaller than <math>t_i+t_f</math>. Thus<br />
<math>TOF-t_0^{(n)} = L_1/v_1^{(n)} + L_2/v_2</math> , n=0, 1, 2,..<br />
Set <math>t_0^{(0)}=0</math> and obtain <math>v_1^{(0)}</math> from the previous formula. From <math>v_1^{(0)}</math> we obtain <math>E_1^{(0)}</math><br />
Set <math>t_0^{(1)}=func( E_1^{(0)} )</math> and repeat the steps until <math>|t_0^{(n+1)} - t_0^{(n+1)}| < tolTOF</math>. With tolTOF=0.1microsecond, only one iteration is needed for convergence.
For indirect instruments featuring an incoming neutron flux having a sufficiently narrow distribution of energies, a linear relationship between t_0 and
the wavelength of the incoming neutron can be established. This relation allows for coding of an algorithm with
faster execution times. For indirect instruments that comply with these conditions, use of [[ModeratorTzeroLinear]] is preferred.
*WIKI*/

//----------------------------------------------------------------------
Expand All @@ -47,8 +55,8 @@ DECLARE_ALGORITHM(ModeratorTzero);
/// Sets documentation strings for this algorithm
void ModeratorTzero::initDocs()
{
setWikiSummary(" Corrects the time of flight of an indirect geometry instrument by a time offset that is dependent on the energy of the neutron after passing through the moderator. ");
setOptionalMessage(" Corrects the time of flight of an indirect geometry instrument by a time offset that is dependent on the energy of the neutron after passing through the moderator.");
setWikiSummary("Corrects the time of flight of an indirect geometry instrument by a time offset that is dependent on the energy of the neutron after passing through the moderator.");
setOptionalMessage("Corrects the time of flight of an indirect geometry instrument by a time offset that is dependent on the energy of the neutron after passing through the moderator.");
}

using namespace Mantid::Kernel;
Expand All @@ -70,7 +78,7 @@ void ModeratorTzero::init()
declareProperty(new WorkspaceProperty<MatrixWorkspace>("InputWorkspace","",Direction::Input,wsValidator), "The name of the input workspace, containing events and/or histogram data, in units of time-of-flight");
//declare the output workspace
declareProperty(new WorkspaceProperty<MatrixWorkspace>("OutputWorkspace","",Direction::Output), "The name of the output workspace");
declareProperty(new Kernel::PropertyWithValue<double>("tolTOF", 0.1, Kernel::Direction::Input),"Tolerance in the calculation of the emission time, in microseconds");
declareProperty(new Kernel::PropertyWithValue<double>("tolTOF", 0.1, Kernel::Direction::Input),"Tolerance in the calculation of the emission time, in microseconds (default:1)");
declareProperty(new Kernel::PropertyWithValue<size_t>("Niter", 1, Kernel::Direction::Input),"Number of iterations (default:1)");

} // end of void ModeratorTzero::init()
Expand All @@ -84,8 +92,10 @@ void ModeratorTzero::exec()

//deltaE-mode (should be "indirect")
std::vector<std::string> Emode=m_instrument->getStringParameter("deltaE-mode");
if(Emode.empty()) throw Exception::InstrumentDefinitionError("Unable to retrieve instrument geometry (direct or indirect) parameter", inputWS->getTitle());
if(Emode[0]!= "indirect") throw Exception::InstrumentDefinitionError("Instrument geometry must be of type indirect.");
if(Emode.empty())
throw Exception::InstrumentDefinitionError("Unable to retrieve instrument geometry (direct or indirect) parameter", inputWS->getTitle());
if(Emode[0]!= "indirect")
throw Exception::InstrumentDefinitionError("Instrument geometry must be of type indirect.");

// extract formula from instrument parameters
std::vector<std::string> t0_formula=m_instrument->getStringParameter("t0_formula");
Expand All @@ -110,11 +120,11 @@ void ModeratorTzero::exec()

const size_t numHists = static_cast<size_t>(inputWS->getNumberHistograms());
Progress prog(this,0.0,1.0,numHists); //report progress of algorithm
PARALLEL_FOR2(inputWS, outputWS);
PARALLEL_FOR2(inputWS, outputWS)
// iterate over the spectra
for (int i=0; i < static_cast<int>(numHists); ++i)
{
PARALLEL_START_INTERUPT_REGION;
PARALLEL_START_INTERUPT_REGION
size_t wsIndex = static_cast<size_t>(i);
double L1=CalculateL1(inputWS, wsIndex); // distance from source to sample or monitor
double t2=CalculateT2(inputWS, wsIndex); // time from sample to detector
Expand All @@ -138,6 +148,9 @@ void ModeratorTzero::exec()
else
tof-=CalculateT0(tof, L1, t2, E1, parser);
outbins[ibin] = tof;
//if(ibin%400==0)
//std::cout<<", "<<tof;
//std::cout<<i<<" "<<ibin<<" "<<tof<<" "<<E1<<std::endl;
}
}
else
Expand All @@ -148,9 +161,9 @@ void ModeratorTzero::exec()
outputWS->dataY(i) = inputWS->dataY(i);
outputWS->dataE(i) = inputWS->dataE(i);
prog.report();
PARALLEL_END_INTERUPT_REGION;
PARALLEL_END_INTERUPT_REGION
}
PARALLEL_CHECK_INTERUPT_REGION;
PARALLEL_CHECK_INTERUPT_REGION

// Copy units
if (inputWS->getAxis(0)->unit().get())
Expand Down Expand Up @@ -205,17 +218,17 @@ void ModeratorTzero::execEvent()

// Loop over the spectra
Progress prog(this,0.0,1.0,numHists); //report progress of algorithm
PARALLEL_FOR1(outputWS);
PARALLEL_FOR1(outputWS)
for (int i = 0; i < static_cast<int>(numHists); ++i)
{
PARALLEL_START_INTERUPT_REGION;
PARALLEL_START_INTERUPT_REGION
size_t wsIndex = static_cast<size_t>(i);
EventList &evlist=outputWS->getEventList(wsIndex);
if( evlist.getNumberEvents() > 0 ) //don't bother with empty lists
{
double L1=CalculateL1(matrixOutputWS, wsIndex); // distance from source to sample or monitor
double t2=CalculateT2(matrixOutputWS, wsIndex); // time from sample to detector
if(t2 >= 0) //t2 < 0 when no detector info is available
if(t2>=0) //t2 < 0 when no detector info is available
{
double tof, E1;
mu::Parser parser;
Expand All @@ -225,10 +238,10 @@ void ModeratorTzero::execEvent()
double min_t0_next=parser.Eval(); // fast neutrons are shifted by min_t0_next, irrespective of tof

// fix the histogram bins
MantidVec & x = evlist.dataX();
for (MantidVec::iterator iter = x.begin(); iter != x.end(); ++iter)
MantidVec &x=evlist.dataX();
for (MantidVec::iterator iter=x.begin(); iter!=x.end(); ++iter)
{
tof = *iter;
tof=*iter;
if(tof<m_t1min+t2)
tof-=min_t0_next;
else
Expand All @@ -251,9 +264,9 @@ void ModeratorTzero::execEvent()
}
}
prog.report();
PARALLEL_END_INTERUPT_REGION;
PARALLEL_END_INTERUPT_REGION
}
PARALLEL_CHECK_INTERUPT_REGION;
PARALLEL_CHECK_INTERUPT_REGION
outputWS->clearMRU(); // Clears the Most Recent Used lists */
} // end of void ModeratorTzero::execEvent()

Expand Down

0 comments on commit 048111b

Please sign in to comment.