API Reference

This section documents the Python API for driving Circulatory Autogen programmatically — the same calls used by the interactive tutorial and by external tools such as the CUFLynx GUI.

The reference pages are generated automatically from the docstrings in src/libcuflynx/, so they stay in sync with the code.

Setup

The repository is circulatory_autogen; the installable package is libcuflynx. Install it and import from the libcuflynx namespace — there is no import path to set up, and no checkout or OpenCOR install is required.

pip install libcuflynx

from libcuflynx.utilities.utility_funcs import get_default_inp_data_dict

The old flat imports were removed in 0.6.0

from param_id.paramID import CVS0DParamID (no libcuflynx. prefix) resolved in 0.4.0 and 0.5.x as a shim that warned. In 0.6.0 it raises ModuleNotFoundError; prefix the import with libcuflynx.. See CHANGELOG.md.

Public API at a glance

The entry points below cover the full generate → simulate → calibrate → analyse pipeline. Each links to its detailed reference page.

Task	Entry point	Page
Build the config dict	get_default_inp_data_dict	Utilities
Generate a model	generate_with_new_architecture	Model generation
Run a simulation	get_simulation_helper / get_simulation_helper_from_inp_data_dict	Simulation
Build observation data	ObsDataCreator	Observation data & utilities
Calibrate parameters	CVS0DParamID	Parameter identification
Sensitivity analysis	SensitivityAnalysis	Sensitivity analysis
Identifiability analysis	IdentifiabilityAnalysis	Identifiability analysis
Run protocols standalone	ProtocolRunner	Protocols

Typical call sequence

from libcuflynx.utilities.utility_funcs import get_default_inp_data_dict
from libcuflynx.scripts.script_generate_with_new_architecture import generate_with_new_architecture
from libcuflynx.solver_wrappers import get_simulation_helper_from_inp_data_dict
from libcuflynx.utilities.obs_data_helpers import ObsDataCreator
from libcuflynx.param_id.paramID import CVS0DParamID

# 1. Configuration (== user_inputs.yaml defaults, then mutate in code)
inp = get_default_inp_data_dict(file_prefix, input_param_file, resources_dir)
inp["sim_time"], inp["pre_time"] = 2, 20.0

# 2. Generate the model
generate_with_new_architecture(inp_data_dict=inp)

# 3. Simulate
sim = get_simulation_helper_from_inp_data_dict(inp)
sim.run()
t, y = sim.get_time(), sim.get_results(variables_to_plot, flatten=True)

# 4. Build observation data
obs = ObsDataCreator()
obs.add_protocol_info(pre_times, sim_times, params_to_change)
obs.add_data_item(entry)
obs_data_dict = obs.get_obs_data_dict()

# 5. Calibrate
pid = CVS0DParamID.init_from_dict(inp)
pid.set_ground_truth_data(obs_data_dict)
pid.set_params_for_id(params_for_id_dict)
pid.run()
pid.simulate_with_best_param_vals()
pid.plot_outputs()

params_for_id_dict is a list of {vessel_name, param_name, min, max, name_for_plotting} entries (the in-memory equivalent of {prefix}_params_for_id.csv); vessel_name may be a single name or a list to share one calibrated parameter across many vessels.