Project

General

Profile

Feature #924

Beam Rotation

Added by Littlefield, Matthew over 9 years ago. Updated almost 9 years ago.

Status:
Rejected
Priority:
Normal
Assignee:
Littlefield, Matthew
Category:
Simulation
Target version:
Start date:
02 March 2012
Due date:
% Done:

0%

Estimated time:
Workflow:
New Issue

Description

I'm not sure if this is already a feature or not but can we have add a rotation to the beam? I.e. we give the position of the beam in the worlds volume and also add a rotation to this as well so the beam can be placed with more flexibility?

#1

Updated by Rogers, Chris over 9 years ago

  • Category set to Simulation
  • Target version set to Future MAUS release

So the basic concept:

In src/map/MapPyBeamMaker we allow several options like "reference", "transverse", etc. So need to add another item to this list called "rotation" that takes a three vector "x", "y", "z". The "rotation" command should apply a rotation to input particle data that

  1. rotates through z by angle beam_def["rotation"]["z"]
  2. rotates through z by angle beam_def["rotation"]["y"]
  3. rotates through z by angle beam_def["rotation"]["x"]
Note that
  • rotation is non-commutative so you have to get the order right.
  • beam_def["reference"] object is actually a MAUS primary particle, so don't edit this unless you want to faff with the whole data structure.
  • there is some code and tests in my branch. I got as far as reading in the new beam command and started to think about doing the rotation.
Tests
  • obviously unit test that the rotation is done correctly
  • also please add an integration test that checks that the rotation is the same as a Geant4 physical rotation. So for example, if you have a geometry like
     ________________
    |                |
    |   BOX B        |
    |________________|    WORLD
    o ->
     ________________
    |                |
    |   BOX A        |
    |________________|
    
    

where o is a particle moving in the direction of the arrow; then apply a rotation to the boxes; then apply the same rotation to the particle; the particle should not intersect the boxes. Maybe you can make something more elegant.

#2

Updated by Rogers, Chris over 9 years ago

  • Assignee changed from Rogers, Chris to Littlefield, Matthew
#3

Updated by Littlefield, Matthew over 9 years ago

Thanks Chris, I'll get on it :-)

#4

Updated by Littlefield, Matthew over 9 years ago

I think I have got it so its pulling the rotation from the configuration defaults. You said that I need to apply this rotation to the input particle data. Could you elaborate a little of this please? I'm not sure how to alter the input particle data?

#5

Updated by Littlefield, Matthew over 9 years ago

Scratch that last comment I was missing didn't see all the code you posted. Would you mind explaining what is going on in this part of your code please?

        if self.rotation == None:
            return hit
        cx = math.cos(self.rotation["x"])
        sx = math.sin(self.rotation["x"])
        cy = math.cos(self.rotation["y"])
        sy = math.sin(self.rotation["y"])
        cz = math.cos(self.rotation["z"])
        sz = math.sin(self.rotation["z"])
        rot_matrix = numpy.matrix([[1., 0.,  0.],
                                  [0., cx, -sx],
                                  [0., sx,  cx],])*\
                     numpy.matrix([[cy, 0.,   sy],
                                  [0., 1.,   0.],
                                  [-sy, 0.,  cy],])*\
                     numpy.matrix([[cz, -sz, 0.],
                                  [sz, cz,  0.],
                                  [0., 0.,  0.],])

I can see that the object hit is

Hit.new_from_dict({'local_weight': 1.0, 'energy': 228.37452951150374, 'pid': -13, 'ey': 0.0, 'ex': 0.0, 'ez': 0.0, 'e_dep': 0.0, 'pz': 197.62770972027104, 'event_number': 0, 'px': 37.46616549724004, 'py': -23.039315276523702, 'charge': 0.0, 'station': 0, 'global_weight': 1.0, 'status': 0, 'proper_time': 0.0, 'path_length': 0.0, 'particleNumber': 0, 'particle_number': 0, 'bx': 0.0, 'by': 0.0, 'bz': 0.0, 'sz': 0.0, 'sy': 0.0, 'sx': 0.0, 'eventNumber': 0, 'mass': 105.6583668, 't': 201756.07158053783, 'y': 52.97936343217816, 'x': 39.98097631577003, 'z': 0.0})

and I take it that this is what we need to add the rotation to?
Cheers :-)

#6

Updated by Rogers, Chris over 9 years ago

I was trying to prepare a transformation matrix to apply using xboa.Hit.Hit.abelian_transformation(...) function. So probably one would perform a transformation in (x,y,z) and the same transformation in (px, py, pz).

#7

Updated by Littlefield, Matthew over 9 years ago

Just to make sure. That transform would be applied to x,px,y,py,z,pz in the hit object I just posted?

#8

Updated by Rogers, Chris over 9 years ago

On the python command line...

import xboa.Hit
help(xboa.Hit.Hit.abelian_transformation)

Returns

Help on method abelian_transformation in module xboa.Hit:

abelian_transformation(self, rotation_list, rotation_matrix, translation_dict={}, origin_dict={}, mass_shell_variable='') unbound xboa.Hit.Hit method
    Perform an abelian transformation about the origin, i.e. V_out - O = R*(V_in-O) + T. 
    Then force E^2 = p^2 + m^2
        rotation_list       = list of variables to be rotated
        rotation_matrix     = matrix R
        translation_dict    = dict of strings from set_variables() to floats. Becomes O
        mass_shell_variable = string from list mass_shell_variables()
        origin_dict         = dict of strings from set_variables() to floats. Becomes t
    e.g. hit.abelian_transformation(['x','px'], array[[1,0.5],[0,1]],{'x':10},'energy') will
    look like a drift space plus a translation

This would be two separate calls, one on ['x', 'y', 'z'] and the other on ['px', 'py', 'pz'], something like

hit.abelian_transformation(['x', 'y', 'z'], rot_matrix)
hit.abelian_transformation(['px', 'py', 'pz'], rot_matrix)
#9

Updated by Littlefield, Matthew over 9 years ago

Ah I see, Thanks very much Chris! :-)

#10

Updated by Rogers, Chris almost 9 years ago

  • Status changed from Open to Rejected

Was this ever implemented?

Also available in: Atom PDF