OpenFTS¶

In order to run OpenFTS as a python module, your python script will look something like

import openfts
fts = openfts.OpenFTS()   # create fts object
...                       # edit the fts object
fts.run()                 # run the fts object
...                       # extract data from fts run

Below is the comprehensive API to interface with this fts object (generated via sphinx.autodoc)

openfts.OpenFTS¶: alias of OpenFTS2

class openfts.OpenFTS2¶

New python interface to OpenFTS C++ backend This interface is designed to be more modular and object oriented

compute_optimal_npw(prime_factors=[2, 3], min_smearing_length=None, cell_lengths=None)¶

Using cell size and minumum species smearing length compute optimal npw using localfts object

Parameters:

prime_factors – list of prime integer factors for allowed npw
min_smearing_length – Minimum smearing length to resolve in box. If unspecified, defaults to use values from the diferent species in box (this option was added to support Gibbs ensemble simulations)
cell_lengths – Cell lengths in each dimension. Currently cell must be orthorhombic. defaults to current simulation cell

get_cell_tensor()¶: query localfts object and return cell tensor (in Rg3 units)

get_cell_volume()¶: query localfts object and return cell volume (in Rg3 units)

get_field_float_type()¶: query localfts object and return field_float_type as string

get_min_smearing_length()¶: query localfts object and return minimum smearing length (in Rg3 units)

get_model_Rgref()¶: query localfts object and return Rg of reference chain (using bref and Nref)

get_model_exchange_to_species_map()¶: query localfts object and return exchange to species field mapping as matrix

get_model_interaction_eigenvalues()¶: query localfts object and return exchange to species field mapping as matrix

get_model_nfields()¶: query localfts object and return number of fields used in model

get_molecule_alpha(imol)¶: query localfts object and return alpha of specified molecule

get_molecule_ncopies(imol)¶: query localfts object and return ncopies of specified molecule

get_runtime_mode()¶: Get the runtime mode of the OpenFTS object.

initialize(logfile=None)¶

Initialize the C++ OpenFTS backend but do not run. Useful to query the fts object, but dont want to run a simulation.

Parameters:: logfile – redirect stdout to logfile

io_get_field_error()¶: Read field_error_total from final step of a completed run

io_get_scalar_operator(key, mode='final')¶

Read a scalar operator from final step of a completed run. Uses utils.get_scalar_operator() function

Parameters:: key – key to read from operators.dat file (or similar). See top line of file for options.

io_get_status()¶: Read the STATUS of a completed run

map_to_particle(**kwargs)¶: Function that takes current FTS values and writes the corresponding particle-based input files. See Driver.ParticleMap for arguements/

run(logfile=None)¶

Run the OpenFTS module. Imports and calls the C++ backend via pybind11.

Parameters:: logfile – redirect stdout to logfile

write_json(filename)¶: Write fts arguments to a json file. Useful for running directly using C++ executable.

openfts.get_scalar_operators(fname, mode)¶

read from fname and convert all data into a dictionary This is the updated version. The old version was mostly a wrapper around get_scalar_operator()

Parameters:

filename – fts object or name of file to read. Typically operators.dat
mode – parsing mode. Options are ‘timeseries’, ‘final’, ‘average’ or ‘smart_average’

Returns:

a dictionary where each operator is stored under a key

Return type:

operators

openfts.get_scalar_operator(fname, key, mode)¶

read from filename to get operator named ‘key’: this function is very similar to openfts.io_get_scalar_operator In fact, I could probably overload this function to either take a openfts object or a filename as the first argument

Parameters:

filename – fts object or name of file to read. Typically operators.dat
key – key to read from operators.dat file (or similar). See top line of file for options.
mode – parsing mode. Options are ‘timeseries’, ‘final’, ‘average’

Returns:

either a scalar or a np.array depending on mode

Return type:

operator