bwheelz36 / particlephasespace Goto Github PK

View Code? Open in Web Editor NEW

12.0 2.0 8.0 34.04 MB

Import, analysis, and export of particle phase space data

Home Page: https://bwheelz36.github.io/ParticlePhaseSpace/

License: GNU General Public License v3.0

Python 99.09% Makefile 0.38% Batchfile 0.53%

accelerator-physics particles phase-space topas topas-mc twiss

particlephasespace's People

Contributors

Stargazers

Watchers

Forkers

benjaminbolling ax3l saadataliphd standardgalactic matthewfeickert 5l1v3r1

particlephasespace's Issues

add a 'generator' data loader?

paper: community guidelines

Community guidelines

There should be clear guidelines for third-parties wishing to:

Contribute to the software
Report issues or problems with the software
Seek support

consider adding 'filter' functionality?

Can already be done by simply manipuldating the pandas dataframe inside PhaseSpace, but could make this more streamlined/automated...

Goal

Develop a common library for dealing with phase space data which can import/export from multiple codes in an extensible manner.

phase space format

I'm thinking of using pandas with the below columns.
one potential limitation is that pandas is not good at handling data larger than available memory...

x
y
z
px
py
pz
t
particle_type (use geant)
weight
anything else?

functionality

different export options (cst, comsol, etc.)
filter by particles to generate electron only, proton only etc. phase spaces
project particles forward/back
basic analytics

update docstrings for print_energy_stats and print_twiss

add functionality to 'compress' the phase space?

as per leo's code in https://github.com/lesnat/p2sat

calculate BetaX, BetaY, BetaZ instead of abs(Beta)

make sure data exporters respect units

pandas data setter doesnt work on individual columns

as also noted here when an individual column of a data frame is called the setter method is never activated.
This is very annoying. It would appear the options are:

create a new instance of Dataframe with the set_item method overwritten
document the acceptable ways to interact with PhaseSpace.

fix docstring for plot_particle_positions_hist_2D

integrate units

At read in

update all data loaders
update docs for writing new data loaders

set_units

still not convinced trace space scatter and histogram plots are equivalent..

harmonise function defs between plot_histogram methods

Update link to allowed columns in new exporter docs

should units be hard coded in data exporters, or be flexible?

for all dunder methods, check that 'other' is correct type

paper: Mention (if applicable) a representative set of past or ongoing research projects using the software and recent scholarly publications enabled by it.

make sure units are preserved whenever new instances are created (and maybe add some tests)

what direction is the beam moving in

A lot of the plotting functions assume we have a phase space scored on a plane moving primarily in the z direction.
there is no reason to suppose this is always the case.
so what can we do about this....

where possible don't make this assumption.
the only place it's painful is in the 'project particles' code. Even there, I think it is possible to generalise...

possible issue with `plot_transverse_trace_space_intensity`

compare the below two cases.
it seems like the bounds are not being properly used to form the image?

PS.plot_transverse_trace_space_intensity(xlim=[-1,1], ylim=[-.004, .004], plot_twiss_ellipse=False)
PS.plot_transverse_trace_space_intensity()

add `self.reset_phase_space()` to project method

Topas data loader occasionally returns NaN for pz

in cases where pz <<< P, it seems that due to machine precision error pz can end up being NaN.

To fix this, we can detect such cases and set pz = 0 with a warning.

add csv exporter

class CSV_Exporter(_DataExportersBase):
    '''
    Export data to a csv format, in particular one which can be read by p2sat read text
    # weight          x (um)          y (um)          z (um)          px (MeV/c)      py (MeV/c)      pz (MeV/c)      t (fs)
    '''

    def _define_required_columns(self):
        self._required_columns = ['x', 'y', 'z', 'px', 'py', 'pz', 'time', 'weight']

    def _export_data(self):

        data_string = ['weight', 'x [um]', 'y [um]', 'z [um]', 'px [MeV/c]', 'py [MeV/c]', 'pz [MeV/c]', 'time [fs]']
        self._PS.ps_data[data_string].to_csv(self._output_location / self._output_name, index=False, header=False)

    def _set_expected_units(self):
        self._expected_units = ParticlePhaseSpaceUnits()('p2_sat_UHI')

cannot achieve consistent read in of energy

now, I think that my basic moment to energy formalism is correct.
I also know it worked for the previous SLAC data.
so this implies that actually the momentum calculations are incorrect?

to do

set up doc build
write tests
set up auto testing
protect main branch

are all 2D position plots using equal axes?

specify units in all 'fill' docstrings

add 'grid' to all plots

Expand phase space

Generate N new particles based on the existing phase space.

This will happen in two parts:

Fit some function to the [x,y,z] points (maybe RBF)
generate N points based on this function
Fit another function to momentum as a function of spatial coordinate
Interpolate the momentum at each generated location

double check the behaviour of the private/public facing data frames....

when a user manually changes the public attribute, to the private aspects get updated?

in plot particle positions with weight=True, double check x and y are the correct way around!

docs have two contents headings on front page

project_particles should update time as well

remove param twiss_ellipse from plot beam intensity

remove `NewDataLoader` from main branch

this was only intended as an example not something that should be in the main code base

fix this

operations you make - following any edits to the underling data, you should manually reset_phase_space, because you have quite likely invalidated any previously calculated quantities. reset_phase_space <https://bwheelz36.github.io/ParticlePhaseSpace/code_docs.html#ParticlePhaseSpace._ParticlePhaseSpace.PhaseSpace.reset_phase_space>__

class Load_p2sat_txt(_DataLoadersBase):
    """
    Adapted from the `p2sat <https://github.com/lesnat/p2sat/blob/master/p2sat/datasets/_LoadPhaseSpace.py>`_
    'txt' loader.

    - Note: One difference is that this implementation has a hard coded seperation value ","
    """
    def _check_input_data(self):
        if not Path(self._input_data).is_file():
            raise FileNotFoundError(f'input data file {self._import_data()} does not exist')
        if not self._particle_type:
            raise Exception('particle_type must be specified when readin p2sat_txt data')

    def _import_data(self):
        # Initialize data lists
        w = []
        x, y, z = [], [], []
        px, py, pz = [], [], []
        t = []

        # Open file
        with open(self._input_data, 'r') as f:
            # Loop over lines
            for line in f.readlines():
                # If current line is not a comment, save data
                if line[0] != "#":
                    data = line.split(",")
                    w.append(float(data[0]))
                    x.append(float(data[1]))
                    y.append(float(data[2]))
                    z.append(float(data[3]))
                    px.append(float(data[4]))
                    py.append(float(data[5]))
                    pz.append(float(data[6]))
                    t.append(float(data[7]))


        self.data[self._columns['x']] = x
        self.data[self._columns['y']] = y
        self.data[self._columns['z']] = z
        self.data[self._columns['time']] = t
        self.data[self._columns['weight']] = w

        self.data[self._columns['particle id']] = np.arange(self.data.shape[0])
        self.data[self._columns['particle type']] = particle_cfg.particle_properties[self._particle_type]['pdg_code']

        self.data[self._columns['px']] = px
        self.data[self._columns['py']] = py
        self.data[self._columns['pz']] = pz