Welcome to pySAC’s documentation!¶
PySAC is a python packaged designed to make it easier to analyse and plot output from VAC and SAC in python.
pySAC currently supports routines for VAC and SAC Input / Output as well as 2D plotting.
Contents:
pySAC Guide¶
pySAC is a python package that is designed to enable analysis and plotting of VAC (Verstatile Advection Code) and SAC (Sheffield Advanced Code) as well as smaug, the GPGPU port of SAC.
Installation¶
pySAC is contained inside a DVS version control system called git, with a hosted repository on BitBucket. To obtain the latest version of the source code you can clone the git repository thus:
git clone git@bitbucket.org:swatsheffield/pysac.git
which you can then install using the included python setup script. There are two options at this point, you can install the code permantly or you can install it in ‘develop’ mode which enable you to edit the code in the directory in which you have just cloned it. To install in devlop mode do the following:
python setup.py develop
which, under linux you will need to run as root.
The following python packages are required to use pySAC:
- matplotlib >= 1.1
- h5py
- numpy
- scipy
Basic Usage¶
Once installed pySAC can be used by importing the component you wish to use in a python script or interpreter to open a SAC file you can now do:
myfile = pysac.io.SACfile("mysacfile.out")
myfile now contains various routines to read the data in the file, as we shall see in more detail later.
Input/Output¶
The first role of pySAC is to enable the reading and writing of SAC or VAC files in either the VAC style fortran unformatted binary file type or the newer HDF5 file structures.
The basic components of any VAC/SAC filetype are:
x array: An array containing all coordinates at any point in the domain
y array: An array containing all the computed variables
- header: A header, stored in some form consisting of at:
- filehead - A not so descriptive file string
- iteration number
- physical time
- number of dimensions
- number of constant parameters in the equations
- number of physcical varibles being solved for (in w array)
- nx
- the equation parameters
- names for the varibles in the w array
pySAC implements a ‘VACdata’ class that will read either the VAC fortran type files or HDF5 files containing VAC/SAC data. There is a specification of VACdata to SACdata for SAC files with magnetic field and pertubation and background varibles in the w array.
SACdata implemetes methods to extract primative variables such as temperature and pressure.
File Description¶
Header Contents¶
The header of VAC / SAC files contains the same information irrespective of the file type. When saving a file using pySAC the routines will automatically reformat the header to fit the file. The keys in the header and examples are listed below:
header = {
'filehead': '3Dmhd_33',
'ndim': 3,
'it': 0,
't': 0,
'nw': 13,
'nx': [128,128,128],
'varnames': ['z', 'x', 'y', 'h', 'm1', 'm2', 'm3', 'e', 'b1', 'b2', 'b3',
'eb', 'rhob', 'bg1', 'bg2', 'bg3', 'gamma', 'eta',
'grav1', 'grav2', 'grav3'],
'neqpar': 7,
'eqpar': [1.66666667, 0.0, 1.0, -274.0, 0.0, 0.0, 0.0]
}
Some extra parameters can be added for HDF5 files, namely: ‘filedesc’ which is a file desciption.
‘varnames’ should contain a label for each corrdinate dimension, a label for each varible in the w array, followed by each equation parameter. So
len(header['varnames']) == nx + nw + neqpar
The above example is not consitent with this.
VAC File Description¶
The unformatted FORTRAN binary files have the following format for each time step stored a header of the following form preceeds the data:
‘filehead’ a string describing the file i.e. ‘2D_mhd22’
- ‘params’ a list containing:
iteration number, physical time, ndim, number of equation params, number of vars in w array i.e. [10, 11.2541, 2, 7, 9], [int, float, int, int, int]
nx coord dimensions e.g. [128,128,128]
‘eqpar’ - equation parameters, neqpars floats.
varnames a list of strings nw long. Holds the names for all the w vars.
x array
w array
These are the defualt binary file type that VAC/SAC uses.
WARNING: These files are compiler and machine dependant, they are not portable and should not be used over the far superior HDF5 files.
pySAC Code Reference¶
This contains a very draft API reference for all the pySAC functions and classes.
Contents¶
Input / Output¶
pysac.io Module¶
pysac.io.yt_fields Module¶
A set of dervied fields for yt 2.x which combine pertubation and background components and define magnitudes, as well as calculate things like charaistic speeds.
Note: These use yt 3.x like field naming conventions
pysac.io.gdf_writer Module¶
Routines for the writing of GDF files
Functions¶
| write_field(gdf_file, field[, field_shape, ...]) | Write a field to an existing gdf file. |
| write_field_u(gdf_file, data, field_title, ...) | Write a field to an existing gdf file. |
| create_file(f, simulation_parameters, ...[, ...]) | Do all the structral creation of a gdf file. |
Classes¶
| SimulationParameters(*args, **kwargs) |
Class Inheritance Diagram¶
pysac.io.legacy Module¶
Classes¶
| SACdata(filename[, filetype]) | This adds specifications to VACdata designed for SAC simulations in 2D or 3D with magnetic field. |
| VACdata(filename[, filetype]) | This is a file type independant class that should expose a VACfile or VACHDF5 file so it is transparent to the end user. |
| VACfile(fname[, mode, buf]) | Base input class for VAC Unformatted binary files. |
| VAChdf5(filename[, mode]) | Based on FortranFile has been modified to read VAC / SAC HDF5 files. |
Class Inheritance Diagram¶
pysac.plot Module¶
Functions¶
| fieldlines(v1, v2, seeds[, dS]) | Simple 2D Euler fieldline integrator. |
Classes¶
| FixedCentre([vmin, vmax, clip]) | Normalise with a Fixed Centre If vmin or vmax is not given, they are taken from the input’s minimum and maximum value respectively. |
Class Inheritance Diagram¶
pysac.plot.mayavi_plotting_functions Module¶
This module provides a set of helper functions to make mayavi scenes into nice plots. It also provides routines for the visualisation of flux surfaces.
Functions¶
| set_text(text_property) | Set the text to sane defaults |
| set_cbar_text(cbar[, lut_manager]) | |
| add_axes(ranges[, obj]) | |
| add_cbar_label(cbar, title) | |
| add_colourbar(module, position, position2, title) | |
| draw_surface(surf_poly, cmap[, scalar, ...]) | Draw a mayavi surface from a PolyData object. |
| change_surface_scalars(new_tube, ...[, ...]) | Change the surface scalars of an existing surface object, and change the associated colorbar and text objects. |
pysac.plot.mayavi_seed_streamlines Module¶
This module contains a custom streamlining class derived from the MayaVi2 streamlining class, modified to accept an array of seed points for visulaisation using mayavi.
Warning
The documentation for this class cannot be built on Read The Docs, it is possible to build it locally.
You can use this class thus:
Create a new Streamline instance and add it to a pipeline
>>> from pysac.plot.mayavi_seed_streamline import SeedStreamline
>>> field_lines = SeedStreamline(seed_points = np.array(seeds))
>>> myvectorfield.add_child(field_lines)
Analysis¶
pysac.analysis Module¶
General routines for the analysis of HD and MHD simulations run with SAC
Functions¶
| get_wave_flux(f, pk) | Calculate the wave energy flux from a SACData instance and the kinetic pressure |
| get_wave_flux_yt(ds[, B_to_SI, V_to_SI, ...]) | Calculate the wave energy flux from a yt dataset. |
pysac.analysis.tube3D.tvtk_tube_functions Module¶
This submodule provides routines to generate and analyse “Flux Surfaces” as described in (Mumford et. al. 2014). Flux Surfaces are created by the tracing of a closed loop of fieldlines, from which a surface is reconstructed by creating polygons between the pesudo parallel streamlines.
Functions¶
| move_seeds(seeds, vfield, dt) | Move a list of seeds based on a velocity field. |
| make_circle_seeds(n, r, **domain) | Generate an array of n seeds evenly spaced in a circle at radius r. |
| create_flux_surface(bfield, surf_seeds) | Create a flux surface from an array of seeds and a tvtk vector field. |
| update_flux_surface(surf_seeds, ...) | Update the flux surface streamlines and surface. |
| make_poly_norms(poly_data) | Extract the normal vectors from a PolyData instance (A surface). |
| norms_sanity_check(poly_norms) | Check that the normals are pointing radially outwards. |
| get_surface_vectors(poly_norms, surf_bfield) | Calculate the vector normal, vertically parallel and Azimuthally around |
| interpolate_scalars(image_data, poly_data) | Interpolate a imagedata scalars to a set points in polydata |
| interpolate_vectors(image_data, poly_data) | Interpolate a imagedata vectors to a set points in polydata |
| update_interpolated_vectors(poly_data, ...) | |
| update_interpolated_scalars(poly_data, ...) | |
| get_surface_velocity_comp(...) | |
| get_the_line(bfield, surf_seeds, n) | Generate the vertical line on the surface |
| update_the_line(the_line, surf_seeds, seed, ...) | Updates the TD line at each time step, while making sure the length is fixed |
| get_surface_indexes(surf_poly, the_line) | |
| write_step(file_name, surface, normals, ...) | Write out the surface vectors and velocity compnents. |
| write_flux(file_name, surface, ...) | |
| write_wave_flux(file_name, surface_poly, ...) | |
| read_step(filename) | Read back in a saved surface file |
| get_data(poly_out, name) |
Classes¶
| PolyDataWriter(filename, polydata) | This class allows you to write tvtk polydata objects to a file, with as many or as few associated PointData arrays as you wish. |
Class Inheritance Diagram¶
pysac.analysis.tube3D.process_utils Module¶
This module contains routines to convert from SACData and yt DataSets into tvtk fields via mayavi.
Functions¶
| get_sacdata_mlab(f, cube_slice[, flux]) | Reads in useful variables from a hdf5 file to vtk data structures |
| get_yt_mlab(ds, cube_slice[, flux]) | Reads in useful variables from yt to vtk data structures, converts into SI units before return. |
| process_next_step_yt(ds, cube_slice, bfield, ...) | Update all mayavi sources using a yt dataset, in SI units. |
| process_next_step_sacdata(f, cube_slice, ...) | Update all mayavi sources using a SACData instance |
| yt_to_mlab_vector(ds, xkey, ykey, zkey[, ...]) | Convert three yt keys to a mlab vector field :Parameters: ds : yt dataset The dataset to get the data from. |
pysac.mhs_atmosphere Module¶
Functions¶
| construct_magnetic_field(x, y, z, x0, y0, S, ...) | Construct self similar magnetic field configuration |
| construct_pairwise_field(x, y, z, xi, yi, ...) | Construct self similar magnetic field configuration |
| get_flux_tubes(model, model_pars, coords, ...) | Obtain an array of x,y coordinates and corresponding vertical |
| get_internal_energy(pressure, magp, ...) | Convert pressures to internal energy – this may need revision if an alternative equation of state is adopted. |
| get_logical(model, l_mpi[, l_SI, l_gdf]) | This module assigns the logical options for the model. |
| get_model(model, Nxyz, xyz_SI, scales, ...) | |
| get_parameters(model, l_mpi, logical_pars, size) | |
| get_spruit_hs(filenames, Z, scales, ...) | |
| interpolate_atmosphere(filenames, Z, scales, ...) | This module generates a 1d array for the model plasma preesure, plasma density, temperature and mean molecular weight. |
| mhs_3D_profile(z, pressure_z, rho_z, ...) | Return the vertical profiles for thermal pressure and density in 3D |
| save_SACsources(model, sourcesfile, Fx, Fy, ...) | Save the balancing forces for a SAC model with multiple flux tubes in hdf5 (gdf default) format after collating the data from mpi sub processes if necessary. |
| save_SACvariables(model, filename, rho, Bx, ...) | Save the background variables for a SAC model in hdf5 (gdf default) format after collating the data from mpi sub processes if necessary. |
| save_auxilliary1D(model, auxfile, ...) | Save auxilliary variables for use in plotting background setup in hdf5 (gdf default) format after collating the data from mpi sub processes if necessary. |
| vertical_profile(Zint, Z, pdata_i, rdata_i, ...) | Return the vertical profiles for thermal pressure and density in 1D. |