Skip to content

Latest commit

 

History

History
59 lines (37 loc) · 2.92 KB

README.md

File metadata and controls

59 lines (37 loc) · 2.92 KB
Raw

MC-discrete (very alpha)

Discrete inelastic electron scattering Monte Carlo python implementation

Required libraries

Probably the easiest thing to do is to add a new Python3 environment to conda and then

conda install:

  • dask
  • pandas

conda install -c conda forge:

  • quaternion

Set up

Bacause I moved the python files in separate folders and used namespaces to call them, python needs to know the path we're in:

`export PYTHONPATH=[where the repo is]

MC/doScatter.py calls the rest of the code according to the parameters set in inputFile/input.file. The parametes in the input file are:

  • number of electron per process : num_el

  • incident beam energy: E0

  • electrons are tracked from energy E_0 to a set minimum energy: Emin

  • material with which the electrons interact material; so far only free electron systems are implemented: Al, Si, Cu, Au

  • the tilt of the sample in degrees defined in the same way as in EMsoft: s_tilt

  • scattering mode, mode, can be continuous or discrete (DS)

  • if continuous mode is chosen then three implementations are available for the Bethe_model: classical, JL or explicit

  • two tolerances can be set for the determining the accuracy of the integration tables for the discrete model: tol_E for the accuracy of the trapezoidal integrals and tol_W for the steps of energy loss determining the accuracy in that direction

  • Wc is a free parameter needed for Moller scattering and represents the minimum energy loss allowed for this channel.

  • the maximum number of scatterings can also be set thorough maxScatt (though this tends to be disables untill I figure out the physics of setting this)

  • the output is made of two dataframes in a hdf5 table with columns set here. electron_output contains all the electron parameters of interenst and scatter_output the scattering values.

Tests

tests/testThings.py runs some statistical tests for the stochastic patameters, set up follows tests/inputTest.file

Thing to finish:

  • The set up should only be called once for all the tests in the class.
  • Gryz scatter test
  • Quinn scatter test
  • the set unit situatuation should be moved to a test

Structure

doScatter.py calls scatterMultiEl_cont or scatterMultiEl_DS from multiScattering.py depending on the mode set. These functions will spawn a multiprocessing job for one electron trajectory for the number of electrons set. For every trajectory a new electron instance is spawned which is allowed to suffer instances of scattering untill the conditions for basckscattering (/transmission in the future), absorbtion (or if set max number of scatterings) are met.

At every scattering the parameters set in scatter_output as saved to a queue and finally added to the output dataset. For every electron the parameters set i electron_output are similarly saved to a queue and then to a output dataset.

Current issues

hdf5 locks for largish data (electrons>500000).