fluidsim.util

Utilities for FluidSim

Provides:

`util`	Utilities for the numerical simulations (fluidsim.util.util)
`testing`	Utilities for the unittests (fluidsim.util.testing)
`console`	Console Utilities for FluidSim See fluidsim -h
`scripts`	Utilities for fluidsim scripts (fluidsim.util.scripts)
`mini_oper_modif_resol`	Mini operator to modify the resolution (Fourier)

fluidsim.util.load_sim_for_plot(name_dir=None, merge_missing_params=False, hide_stdout=False)[source]

Create a object Simul from a dir result.

Creating simulation objects with this function should be fast because the state is not initialized with the output file and only a coarse operator is created.

Parameters:

name_dirstr (optional): Name of the directory of the simulation. If nothing is given, we load the data in the current directory. Can be an absolute path, a relative path, or even simply just the name of the directory under $FLUIDSIM_PATH.
merge_missing_paramsbool (optional, default == False): Can be used to load old simulations carried out with an old fluidsim version.
hide_stdoutbool (optional, default == False): If True, without stdout.

fluidsim.util.load_state_phys_file(name_dir=None, t_approx='last', modif_save_params=True, merge_missing_params=False, init_with_initialized_state=True, hide_stdout=False)[source]

Create a simulation from a file.

For large resolution, creating a simulation object with this function can be slow because the state is initialized with the output file.

Parameters:

name_dirstr (optional)

Name of the directory of the simulation. If nothing is given, we load the data in the current directory. Can be an absolute path, a relative path, or even simply just the name of the directory under $FLUIDSIM_PATH.

t_approxnumber (optional)

Approximate time of the file to be loaded.

modif_save_paramsbool (optional, default == True)

If True, the parameters of the simulation are modified before loading:

params.output.HAS_TO_SAVE = False
params.output.ONLINE_PLOT_OK = False

merge_missing_paramsbool (optional, default == False)

Can be used to load old simulations carried out with an old fluidsim version.

init_with_initialized_statebool (optional, default == True)

If True, call sim.output.init_with_initialized_state.

hide_stdoutbool (optional, default == False)

If True, without stdout.

fluidsim.util.load_for_restart(name_dir=None, t_approx='last', merge_missing_params=False)[source]

Load params and Simul for a restart.

>>> params, Simul = load_for_restart(name_dir)

Parameters:

name_dirstr (optional): Name of the directory of the simulation. If nothing is given, we load the data in the current directory. Can be an absolute path, a relative path, or even simply just the name of the directory under $FLUIDSIM_PATH.
t_approxnumber (optional): Approximate time of the file to be loaded.
merge_missing_paramsbool (optional, default == False): Can be used to load old simulations carried out with an old fluidsim version.

fluidsim.util.load_params_simul(path=None, only_mpi_rank0=True): Load the parameters and return a Parameters instance.

fluidsim.util.modif_resolution_from_dir(name_dir=None, t_approx='last', coef_modif_resol=2, PLOT=True)[source]: Save a file with a modified resolution.

fluidsim.util.modif_resolution_from_dir_memory_efficient(name_dir=None, t_approx='last', coef_modif_resol=2)[source]

Save a file with a modified resolution.

Faster and more memory efficient than modif_resolution_from_dir (but not plot).

fluidsim.util.times_start_last_from_path(path)[source]: Return the start and last times from a result directory path.

fluidsim.util.ensure_radians(angle)[source]: Convert strings like “45°” to radians (as float)

fluidsim.util.get_mean_values_from_path(path, tmin=None, tmax=None, use_cache=True, customize=None)

Get a dict of scalar values characterizing the simulation

Parameters:

tmin: float: Minimum time
tmax: float: Maximum time
use_cache: bool: If True, return the cached result
customize: callable: If not None, called as customize(result, self.sim) to modify the returned dict.

Examples

def customize(result, sim):
    result["Rb"] = float(sim.params.short_name_type_run.split("_Rb")[-1])
get_mean_values_from_path(path, customize=customize)

fluidsim.util.get_dataframe_from_paths(paths, tmin=None, tmax=None, use_cache=True, customize=None)

Produce a dataframe from a set of simulations.

Uses sim.output.get_mean_values

fluidsim.util.get_last_estimated_remaining_duration(path)[source]: Get last estimated remaining duration written in stdout.txt

fluidsim.util.open_patient(path, *args, time_wait_total=200, time_wait_once=2, class_file=<class 'h5py._hl.files.File'>, **kwargs)[source]

Open a hdf5 type file in a “patient” way.

If the file is already opened by another process (errno==11 for hdf5), the error is caught and we retry later.

Parameters:

time_wait_totalnumber (optional): Time to wait before raising the error.
time_wait_oncenumber (optional): Time between attempts.
class_filetype (optional): Class of the file (default h5py.File, but could be h5netcdf.File).