KevinCWLi
commented
4 years ago Greetings K600
I've been assigned the project providing an example of the new developments of the k600analyser that have been used for the PhD thesis of KCW Li. The following repos have been updated to be consistent with the submitted results in the PhD thesis:
PR166_2010 (not used for the PhD work, but has been updated with the latest version) PR166_2011 PR194 PR195_2017 PR240 PR251
A good example to look at with relatively-little hardcoding is PR194: 12C(a,a')12C at 0 deg, Ebeam = 160 MeV. In this email, I've focused on the SilverBulletRaytrace development, however the TotalLineshapeCorrection() and X1mapping() functions could possibly also be of interest to future works. I've attempted to break this email up with comments to try and keep sections clear.
------------------------------------------------
The pre-processor variable shown below is how the SilverBulletRaytrace class is implemented. For each wireplane, a SilverBulletRaytrace object should be defined. Since this type of analysis is typically only applicable for the first horizontal wireplane (X1), only a single SilverBulletRaytrace object has been implemented in the example. The green code is where the code is made to work with functions that are not currently defined in the master branch. They can be implemented however people see fit and each green section has been (hopefully) described.
define _SILVERBULLETRAYTRACE
ifdef _SILVERBULLETRAYTRACE
SilverBulletRaytrace *silverBulletRaytrace_X1 = new SilverBulletRaytrace();
endif
------------------------------------------------
The branch is implemented as:
ifdef _SILVERBULLETRAYTRACE
gROOT->ProcessLine("#include \"SilverBulletRaytrace.h\"");
gROOT->ProcessLine(".L SilverBulletRaytrace.c+");
t1->Branch("SilverBulletRaytrace_X1","SilverBulletRaytrace", &silverBulletRaytrace_X1);endif
------------------------------------------------
This takes the raw drift distances (i.e. without assumption of which wires are inverted) and executes the SBR algorithm of searching through all permutations (under some conditions).
Clearly, it is explicit that only sets of wires with >=3 and <=12 are considered as valid wireplane events... outside of this range, a focal-plane event within the SilverBulletRaytrace class is not created.
ifdef _SILVERBULLETRAYTRACE
if(X1hits_dt>=3 && X1hits_dt<=12)
{
silverBulletRaytrace_X1->Raytrace(X1.dist, X1.wire, X1hits_dt);
}endif
------------------------------------------------
For each "Raytrace" function call (see above), a total of N positions, returned by GetNAlternativeEvents(), is created.
Each of these positions can then undergo lineshape correction and the "SilverBulletRaytrace" object, silverBulletRaytrace_X1, then stores each corresponding lineshape-corrected position.
Or in other words, there is a vector of focal-plane positions determined by the Raytrace() function of SilverBulletRaytrace. This class also stores each corresponding line-shape corrected position.
The "TotalLineshapeCorrection" algorithm is not currently implemented in the master branch.
It is used for lineshape correction and could be implemented at the same time... but the point is that here is where the lineshape correction should be done.
ifdef _SILVERBULLETRAYTRACE
for(int i=0; i<silverBulletRaytrace_X1->GetNAlternativeEvents(); i++)
{
double xPosition = silverBulletRaytrace_X1->GetXPosition(i);
TotalLineshapeCorrection(correctionParameters, &xPosition);
//TotalLineshapeCorrection(X1pos, t_Y1, thetaSCAT, &xPosition);
silverBulletRaytrace_X1->SetXPosition_TLC(i, xPosition);
}endif
------------------------------------------------
Similarly to the lineshape correction above, the lineshape-corrected positions can be mapped (i.e. offset, for simpler cases) and the SilverBulletRaytrace class also stores each mapped position.
ifdef _SILVERBULLETRAYTRACE
extern bool X1MappingDefined_SBR;
for(int i=0; i<silverBulletRaytrace_X1->GetNAlternativeEvents(); i++)
{
if(X1MappingDefined_SBR)
{
double correctedXPosition = X1Mapping(silverBulletRaytrace_X1->GetXPosition_TLC(i));
silverBulletRaytrace_X1->SetXPosition_TLC_mapped(i, correctedXPosition);
}
}endif
------------------------------------------------
Similarly for the preceding lineshape-correction and peak-position mapping sections, the SilverBulletRaytrace object also stores the excitation-energy associated with each lineshape-corrected and peak-position mapped focal-plane position.
The CalcEx() function is not implemented in the master branch but is provided in a modified FocalPlane.c file.
It obviously doesn't need to be used (unless you want to), any excitation-energy mapping can be implemented here (unless you want to do it after sorting, directly from the TTree).
ifdef _SILVERBULLETRAYTRACE
for(int i=0; i<silverBulletRaytrace_X1->GetNAlternativeEvents(); i++)
{
if(momentumCalibrationRead_SBR)
{
double excitationEnergy = CalcEx(&m_SBR[0], &T_SBR[0], &E_SBR[0], &p_SBR[0], silverBulletRaytrace_X1->GetXPosition_TLC_mapped(i), polarScatteringAngles_SBR[2], momentumCalPars_SBR);
silverBulletRaytrace_X1->SetExcitationEnergy(i, excitationEnergy);
}
}
silverBulletRaytrace_X1->DetermineGoodSbrEvent();endif
------------------------------------------------
At the end of the "main_event" routine, the following class function is called to clear the objects.
ifdef _SILVERBULLETRAYTRACE
silverBulletRaytrace_X1->ClearEvent();endif
[http://cdn.sun.ac.za/100/ProductionFooter.jpg]http://www.sun.ac.za/english/about-us/strategic-documents
The integrity and confidentiality of this email are governed by these terms. Disclaimerhttp://www.sun.ac.za/emaildisclaimer Die integriteit en vertroulikheid van hierdie e-pos word deur die volgende bepalings bere?l. Vrywaringsklousulehttp://www.sun.ac.za/emaildisclaimer
padsley
Putting my project-manager hat on for a moment.
We have made fairly extensive improvements to e.g. the gamma-ray processing codes. These have been made in sub-branches of the K600 analyser.
I would like to implement these improvements in the master branch.
Please let me know which branches contain the improvements by creating an issue on the Github repo and tagging the issue in a reply to this issue. I shall probably implement the developments manually because there are a lot of changes made independently and that's probably the quickest way.
@KevinCWLi - I would like to take the opportunity to make sure that the new ray-tracing algorithm is either implemented instead of the old one or in addition (giving it as another option). Could you like me have a working version of the new raytracing algorithm so that it can be implemented?