Reference·Hand-written behaviour attached to the generated model classes.

dynamics_runtime

behaviour.dynamics_runtime

What a Dynamics does, on every Dynamics however it was built.

Attached through DynamicsBehaviour, so a model loaded through LinkML, validated through Pydantic or resolved onto an edge answers the same questions as one constructed through :mod:tvbo.classes.dynamics. There is no runtime subclass to promote a record into: the methods are on the generated class itself.

Split out of DynamicsBehaviour only to keep one file readable; the mixin discovery in hatch_build attaches classes whose name ends in Behaviour, so this one adds no second attachment point. :mod:tvbo.classes.dynamics remains the public import location for the class itself.

Attributes

Name	Description
analysis
data_types
expression
model_helpers
nx
ontology
owlready2
perturbation
plt
query
report
templater
templates
tvbo_datamodel
utils
yaml_loader

Classes

Name	Description
DynamicsRuntime	Construction, code generation, simulation, plotting and reporting for a model.

DynamicsRuntime

behaviour.dynamics_runtime.DynamicsRuntime()

Construction, code generation, simulation, plotting and reporting for a model.

Attributes

Name	Description
components	Alias for modes — sub-dynamics contained in this model.
metadata	The underlying datamodel instance holding the schema fields (this object).
ontology	The ontology class matching this model’s name, or None if it is not a known neural mass model.

Methods

Name	Description
add_coupling_input	Add a coupling input (a network-supplied term) to the model.
add_derived_parameter	Add a derived parameter (computed from other parameters) to the model.
add_derived_variable	Add a derived (algebraic) variable to the model.
add_function	Add a reusable function definition to the model.
add_output	Add an output variable. Creates a derived_variable and adds its name to output list.
add_parameter	Add a parameter to the model.
add_state_variable	Add a state variable (with its differential equation) to the model.
add_stimulus	Attach a stimulus to the model.
animate	Animate by sweeping one parameter through values.
copy	Return a deep copy of this experiment.
db_overview	Return a pandas DataFrame summarising the Dynamics database.
display_markdown	Render generated code as an IPython Markdown code block.
execute	Generate and execute the model code, returning a runnable object.
fill_in_equations	Substitute parameter values (and any overrides) into every equation.
find_periodic_orbits	Find sibling periodic-orbit output files for a run.
from_datamodel	Create from a datamodel Dynamics instance by copying its already-normalized state (avoids _as_dict re-init crash on inlined_as_dict fields).
from_db	Load a Dynamics model by name from the tvbo database.
from_file	Load a model from a YAML/JSON specification file on disk.
from_ontology	Create a model populated from an ontology class.
from_platform	Load a dynamics model from the tvbo platform API.
from_pydantic	Create a Dynamics from a tvbopydantic.Dynamics (or dict-like).
from_pyrates	Load a Dynamics model from a PyRates YAML template file.
from_string	Load a model from a YAML specification string.
generate_report	Render a human-readable report of the model.
get_dependency_tree	Build the equation dependency graph for this model.
get_initial_values	Build the initial state vector for a simulation.
get_run_filename	Build a deterministic cache filename for a run in the temp directory.
list_db	List available models in the tvbo database.
list_platform_models	List available dynamics models on the tvbo platform.
parameter_table	Return a pandas DataFrame of the model’s parameters.
plot	Plot trajectories of this dynamics in 1D, 2D, or 3D.
plot_bifurcation_timeseries	Plot a bifurcation diagram alongside representative time series.
plot_dependency_tree	Plot the model’s equation dependency graph.
plot_ontology	Plot this model’s ontology graph.
render	Unified entry point for rendering the model in any output format.
render_code	Generate backend source code for this model.
render_equation	Render a model element’s equation to a string.
render_equation_cse	Common-subexpression-eliminated variant of :meth:render_equation.
run	Generate, execute, and integrate the model, returning its output.
save_model_metadata	Serialize the model metadata to a YAML file.
save_python_class	Write the model as a standalone TVB Python class file.
save_report	Generate the model report and write it to a directory.
search_ontology	Search this model’s ontology subtree for a term.
symbolic_rhs	The parsed right-hand side of one of this model’s elements.
to_lems	Build a LEMS model for this local neural mass model.
to_pydantic	Return a tvbopydantic.Dynamics validated instance for this model.
to_yaml	Export the model to YAML format.
update_parameters_from_equations	Scan all equations and add any free symbols as parameters (default value if missing).

add_coupling_input

behaviour.dynamics_runtime.DynamicsRuntime.add_coupling_input(
    name,
    description=None,
    unit=None,
    dimension=1,
    keys=None,
)

Add a coupling input (a network-supplied term) to the model.

Any existing parameter with the same name is removed so the name resolves to the coupling input.

Parameters

Name	Type	Description	Default
name	str	Coupling-input name (also its dict key).	required
description	str | None	Human-readable description.	None
unit	str | None	Accepted for backward compatibility; not currently stored on the coupling input.	None
dimension	int	Number of components the input carries.	1
keys	list[str] | None	Optional sub-keys addressed by the coupling input.	None

Returns

Name	Type	Description
		self, to allow fluent chaining.

add_derived_parameter

behaviour.dynamics_runtime.DynamicsRuntime.add_derived_parameter(
    name,
    expression=None,
    *,
    unit=None,
    description=None,
    symbol=None,
)

Add a derived parameter (computed from other parameters) to the model.

Parameters

Name	Type	Description	Default
name	str	Derived-parameter name (also its dict key).	required
expression		RHS expression; accepts a string, sympy.Eq/Expr, or Equation.	None
unit	str | None	Physical unit.	None
description	str | None	Human-readable description.	None
symbol	str | None	Display symbol.	None

Returns

Name	Type	Description
		self, to allow fluent chaining.

add_derived_variable

behaviour.dynamics_runtime.DynamicsRuntime.add_derived_variable(
    name,
    expression=None,
    *,
    conditionals=None,
    unit=None,
    description=None,
    symbol=None,
)

Add a derived (algebraic) variable to the model.

Parameters

Name	Type	Description	Default
name	str	Derived-variable name (also its dict key).	required
expression		RHS expression; accepts a string, sympy.Eq/Expr, or Equation.	None
conditionals	list[tuple[object, object]] | None	Optional list of (expression, condition) pairs defining a piecewise/conditional variable.	None
unit	str | None	Physical unit.	None
description	str | None	Human-readable description.	None
symbol	str | None	Display symbol.	None

Returns

Name	Type	Description
		self, to allow fluent chaining.

add_function

behaviour.dynamics_runtime.DynamicsRuntime.add_function(
    name,
    expression=None,
    *,
    arguments=(),
    description=None,
    definition=None,
)

Add a reusable function definition to the model.

Parameters

Name	Type	Description	Default
name	str	Function name (also its dict key).	required
expression		Function body; accepts a string, sympy.Eq/Expr, or Equation.	None
arguments		Argument names as a sequence, or a mapping of name to Parameter.	()
description	str | None	Human-readable description.	None
definition	str | None	Formal definition or ontology reference.	None

Returns

Name	Type	Description
		self, to allow fluent chaining.

add_output

behaviour.dynamics_runtime.DynamicsRuntime.add_output(
    name,
    expression=None,
    *,
    unit=None,
    description=None,
)

Add an output variable. Creates a derived_variable and adds its name to output list.

add_parameter

behaviour.dynamics_runtime.DynamicsRuntime.add_parameter(
    name,
    value=None,
    unit=None,
    description=None,
    domain=None,
    definition=None,
    symbol=None,
)

Add a parameter to the model.

Parameters

Name	Type	Description	Default
name	str	Parameter name (also its dict key).	required
value	float | None	Numeric default value.	None
unit	str | None	Physical unit.	None
description	str | None	Human-readable description.	None
domain		Valid range as a Range, (lo, hi[, step]) tuple, or dict.	None
definition	str | None	Formal definition or ontology reference.	None
symbol	str | None	Display symbol.	None

Returns

Name	Type	Description
		self, to allow fluent chaining.

add_state_variable

behaviour.dynamics_runtime.DynamicsRuntime.add_state_variable(
    name,
    equation=None,
    *,
    description=None,
    domain=None,
    boundaries=None,
    initial_value=0.1,
    unit=None,
    coupling_variable=False,
    stimulation_variable=None,
    symbol=None,
)

Add a state variable (with its differential equation) to the model.

Any free symbols in equation that are not yet known are auto-registered as parameters. A legacy boundaries clamp is folded into domain (with the descriptive range preserved as the sampling distribution).

Parameters

Name	Type	Description	Default
name	str	State-variable name (also its dict key).	required
equation		RHS of its time-derivative equation; accepts a string, sympy.Eq/Expr, or Equation.	None
description	str | None	Human-readable description.	None
domain		Valid/sampling range as a Range, tuple, or dict.	None
boundaries		Legacy hard-clamp range, folded into domain.	None
initial_value	float | None	Default initial condition.	0.1
unit	str | None	Physical unit.	None
coupling_variable	bool	Mark this variable as observed for network coupling.	False
stimulation_variable	bool | None	Mark this variable as a stimulation target.	None
symbol	str | None	Display symbol.	None

Returns

Name	Type	Description
		self, to allow fluent chaining.

add_stimulus

behaviour.dynamics_runtime.DynamicsRuntime.add_stimulus(
    stimulus,
    as_derived_variable=True,
)

Attach a stimulus to the model.

Warns if no state variable is marked as a stimulation target. Depending on as_derived_variable, the stimulus is either stored on self.stimulus or lowered into a stim_t derived variable plus suffixed stimulus parameters.

Parameters

Name	Type	Description	Default
stimulus		A Stimulus to apply.	required
as_derived_variable		If True, inline the stimulus as a stim_t derived variable; if False, store the Stimulus object directly.	True

animate

behaviour.dynamics_runtime.DynamicsRuntime.animate(
    parameter,
    values,
    *dims,
    **kwargs,
)

Animate by sweeping one parameter through values.

See :func:tvbo.plot.dynamics.animate_dynamics for parameters. Returns a :class:matplotlib.animation.FuncAnimation.

copy

behaviour.dynamics_runtime.DynamicsRuntime.copy(**overrides)

Return a deep copy of this experiment.

Use keyword overrides to set attributes on the returned copy.

Errors are not swallowed; if a field can’t be copied, an exception is raised.

db_overview

behaviour.dynamics_runtime.DynamicsRuntime.db_overview(model_type=None)

Return a pandas DataFrame summarising the Dynamics database.

Columns: name, model_type, system_type, description.

Parameters

model_type : str, optional If given, only show models of that category.

Examples:

Dynamics.db_overview() Dynamics.db_overview(model_type=‘neural_mass’)

display_markdown

behaviour.dynamics_runtime.DynamicsRuntime.display_markdown(
    format='tvb',
    **kwargs,
)

Render generated code as an IPython Markdown code block.

Parameters

Name	Type	Description	Default
format		Backend passed to render_code.	'tvb'
**kwargs		Forwarded to render_code.	{}

Returns

Name	Type	Description
		An IPython.display.Markdown object wrapping the generated code.

execute

behaviour.dynamics_runtime.DynamicsRuntime.execute(format='tvb', **kwargs)

Generate and execute the model code, returning a runnable object.

Dispatches on format: builds a configured TVB model instance, a tvboptim dynamics instance, a compiled C module (sympy2c), a bifurcation/continuation run, or a plain dfun callable.

Every code format resolves its binding through entry_point_name, the same declaration the templates emit against, so any backend that renders Python hands back a usable object for a custom JAX, NumPy or SciPy workflow rather than failing on a name the caller had to guess.

Parameters

Name	Type	Description	Default
format		Backend to execute, e.g. "tvb", "tvboptim", "c", "bifurcation-auto7p", or a code format yielding a dfun.	'tvb'
**kwargs		Constructor/runtime arguments forwarded to the executed code.	{}

Returns

Name	Type	Description
		The executed object, whose type depends on format.

Raises

Name	Type	Description
	ValueError	If format declares no entry point, so nothing can be re-entered from the rendered source.

fill_in_equations

behaviour.dynamics_runtime.DynamicsRuntime.fill_in_equations(**kwargs)

Substitute parameter values (and any overrides) into every equation.

Parameter symbols are replaced with their numeric values, then any **kwargs overrides are applied, and finally all coupling inputs are forced to 0 (for fixed-point / equilibrium analysis) — so a kwargs entry named after a coupling input is overridden by that 0.

Parameters

Name	Type	Description	Default
**kwargs		Additional symbol-name to value substitutions.	{}

Returns

Name	Type	Description
		The list of equations with substitutions applied.

find_periodic_orbits

behaviour.dynamics_runtime.DynamicsRuntime.find_periodic_orbits(f)

Find sibling periodic-orbit output files for a run.

Parameters

Name	Type	Description	Default
f		Path to the base run file whose periodic-orbit companions (<base>_po*) are searched for in the same directory.	required

Returns

Name	Type	Description
		The list of matching periodic-orbit file paths.

from_datamodel

behaviour.dynamics_runtime.DynamicsRuntime.from_datamodel(model_meta)

Create from a datamodel Dynamics instance by copying its already-normalized state (avoids _as_dict re-init crash on inlined_as_dict fields).

from_db

behaviour.dynamics_runtime.DynamicsRuntime.from_db(name)

Load a Dynamics model by name from the tvbo database.

from_file

behaviour.dynamics_runtime.DynamicsRuntime.from_file(path)

Load a model from a YAML/JSON specification file on disk.

Parameters

Name	Type	Description	Default
path	str | os.PathLike	Path to a TVBO model specification file.	required

Returns

Name	Type	Description
	Dynamics	The instance parsed from the file.

from_ontology

behaviour.dynamics_runtime.DynamicsRuntime.from_ontology(ontoclass, **kwargs)

Create a model populated from an ontology class.

Parameters

Name	Type	Description	Default
ontoclass	owlready2.ThingClass | str	An owlready2 model class, or a label string that is resolved against the NeuralMassModel ontology branch.	required
**kwargs		Extra fields forwarded to the constructor and ontology population.	{}

Returns

Name	Type	Description
		A populated instance.

from_platform

behaviour.dynamics_runtime.DynamicsRuntime.from_platform(
    name,
    base_url=TVBO_PLATFORM_URL,
)

Load a dynamics model from the tvbo platform API.

Fetches the full LinkML-valid YAML definition from the platform and constructs a Dynamics instance.

Parameters

name : str Model name (e.g., “JansenRit”, “ReducedWongWang”). base_url : str Platform base URL.

Returns:

Dynamics Dynamics instance loaded from the platform.

from_pydantic

behaviour.dynamics_runtime.DynamicsRuntime.from_pydantic(pyd_obj)

Create a Dynamics from a tvbopydantic.Dynamics (or dict-like).

from_pyrates

behaviour.dynamics_runtime.DynamicsRuntime.from_pyrates(path, operator_key=None)

Load a Dynamics model from a PyRates YAML template file.

Parameters

path : str Path to PyRates YAML file. operator_key : str, optional Name of the specific OperatorTemplate to load (without _op suffix). If None, loads the first OperatorTemplate found. Use SimulationExperiment.from_pyrates() to load all operators.

Returns:

Dynamics New Dynamics instance populated from the PyRates template.

Example:

model = Dynamics.from_pyrates(“jansen_rit.yaml”) # Load specific operator from multi-operator file tsodyks = Dynamics.from_pyrates(“synaptic_plasticity.yaml”, operator_key=“tsodyks”)

from_string

behaviour.dynamics_runtime.DynamicsRuntime.from_string(str)

Load a model from a YAML specification string.

Parameters

Name	Type	Description	Default
str	str	A YAML document describing the model.	required

Returns

Name	Type	Description
	Dynamics	The instance parsed from the string.

generate_report

behaviour.dynamics_runtime.DynamicsRuntime.generate_report(
    format='markdown',
    template_name='tvbo-report-model',
    outputfile=None,
    derivative_notation='dot',
    baseline=None,
    citeformat=None,
)

Render a human-readable report of the model.

Reads the model and does not modify it, for the same reason as render_code: normalisation belongs to construction, and repeating it here made a report a command as well as a query.

Refreshes metadata and renders the Markdown report template; the result is optionally written to outputfile (as Markdown or, for format="pdf", a PDF).

Parameters

Name	Type	Description	Default
format		"markdown"/"md" or "pdf".	'markdown'
template_name		Base name of the report Mako template.	'tvbo-report-model'
outputfile		If given, path the report is written to.	None
derivative_notation	str	Notation for time derivatives, e.g. "dot".	'dot'
baseline		Another Dynamics to diff against. When given, the report lists only the state variables, parameters, derived variables and couplings that are new or changed relative to it (a “relative to” note replaces the shared rows) — e.g. a controlled variant shown against its uncontrolled base without repeating every shared term.	None
citeformat		How references are emitted. Default (None) renders a formatted References section at the end (a standalone report). "quarto" instead emits inline @key citations in the fulltext and omits the list, so the report can be embedded in a Quarto document whose own bibliography: resolves the citations into one bibliography.	None

Returns

Name	Type	Description
		The rendered Markdown report string.

Raises

Name	Type	Description
	ValueError	If format is not one of markdown, md, or pdf.

get_dependency_tree

behaviour.dynamics_runtime.DynamicsRuntime.get_dependency_tree(
    ontomapping=False,
    include_state_equations=False,
)

Build the equation dependency graph for this model.

Nodes are the model’s symbols; each edge points from a dependency to the quantity whose equation uses it (dependencies → dependents). State equations are excluded by default to avoid cycles in discrete systems.

Parameters

Name	Type	Description	Default
ontomapping		If True, also build an ontology-class version of the graph and the symbol↔︎ontology-class mappings.	False
include_state_equations		If True, include state equations in the graph.	False

Returns

Name	Type	Description
		The dependency graph, or — when ontomapping is True — the
		tuple (graph, ontology_graph, symbol_to_onto, onto_to_symbol).

get_initial_values

behaviour.dynamics_runtime.DynamicsRuntime.get_initial_values(
    default=0.1,
    random=False,
    N=1,
    **kwargs,
)

Build the initial state vector for a simulation.

If any state variable defines a distribution (or random=True), initial values are sampled from it (Gaussian or uniform over the finite domain bounds); otherwise each variable’s initial_value (or default) is used.

Parameters

Name	Type	Description	Default
default		Fallback value for variables without an initial value.	0.1
random		Deprecated flag to sample from each variable’s domain.	False
N		Number of samples per state variable.	1
**kwargs		Ignored extra arguments.	{}

Returns

Name	Type	Description
		A NumPy array of initial values. When sampling from a distribution
		(or random=True) the shape is (n_state_variables, N); otherwise
		it is 1-D with one entry per state variable.

get_run_filename

behaviour.dynamics_runtime.DynamicsRuntime.get_run_filename(format, **kwargs)

Build a deterministic cache filename for a run in the temp directory.

Non-identifying keyword arguments (e.g. filename, force, verbose) are dropped and the rest are sorted so the same run maps to the same path.

Parameters

Name	Type	Description	Default
format		Backend format string included in the filename.	required
**kwargs		Run parameters encoded into the filename.	{}

Returns

Name	Type	Description
		The cache-file path (without extension) inside the temp directory.

list_db

behaviour.dynamics_runtime.DynamicsRuntime.list_db(model_type=None)

List available models in the tvbo database.

Parameters

model_type : str, optional Filter by model category. Valid values: mean_field, neural_mass, phase_oscillator, phenomenological, spiking, generic, field.

Examples:

Dynamics.list_db() # all models Dynamics.list_db(model_type=‘mean_field’) # mean-field only Dynamics.list_db(model_type=‘spiking’) # spiking models

list_platform_models

behaviour.dynamics_runtime.DynamicsRuntime.list_platform_models(
    base_url=TVBO_PLATFORM_URL,
    **filters,
)

List available dynamics models on the tvbo platform.

Parameters

base_url : str Platform base URL. **filters Filtering parameters (e.g., system_type=“continuous”).

Returns:

list[dict] List of model summaries.

parameter_table

behaviour.dynamics_runtime.DynamicsRuntime.parameter_table()

Return a pandas DataFrame of the model’s parameters.

Returns

Name	Type	Description
		A DataFrame with Parameter, Value, and Description columns.

plot

behaviour.dynamics_runtime.DynamicsRuntime.plot(*dims, **kwargs)

Plot trajectories of this dynamics in 1D, 2D, or 3D.

See :func:tvbo.plot.dynamics.plot_dynamics for parameters.

plot_bifurcation_timeseries

behaviour.dynamics_runtime.DynamicsRuntime.plot_bifurcation_timeseries(
    ICS,
    VOI,
    n_runs=2,
    t=np.arange(0, 500, 0.1),
    offset=2,
    ax1=None,
    ax2=None,
    **kwargs,
)

Plot a bifurcation diagram alongside representative time series.

Builds two linked panels — a bifurcation diagram over ICS and time series of VOI sampled at several parameter values — and either returns the combined figure or draws into the supplied axes.

Parameters

Name	Type	Description	Default
ICS		Name of the continuation/bifurcation parameter to vary.	required
VOI		Variable of interest to plot.	required
n_runs		Number of parameter values sampled for the time-series panel.	2
t		Time vector for the time-series simulations.	np.arange(0, 500, 0.1)
offset		Vertical offset between successive time-series traces.	2
ax1		Axis for the bifurcation panel; a new layout is made if omitted.	None
ax2		Axis for the time-series panel.	None
**kwargs		Forwarded to the bifurcation run.	{}

Returns

Name	Type	Description
		The combined figure when axes are not supplied, otherwise None.

plot_dependency_tree

behaviour.dynamics_runtime.DynamicsRuntime.plot_dependency_tree(
    ax=None,
    edgecolor='#426665',
    color_nodes_by=None,
    pos='graphviz',
    edgekwargs=None,
    **kwargs,
)

Plot the model’s equation dependency graph.

Parameters

Name	Type	Description	Default
ax		Existing matplotlib axis to draw into; if omitted, a new figure is created and returned.	None
edgecolor		Node edge color.	'#426665'
color_nodes_by		Ontology attribute used to color nodes by category.	None
pos		Node layout, "graphviz" (hierarchical) or otherwise a Kamada–Kawai layout.	'graphviz'
edgekwargs		Extra keyword arguments for edge drawing.	None
**kwargs		Forwarded to the node-drawing helper.	{}

Returns

Name	Type	Description
		The created figure when ax was not supplied, otherwise None.

plot_ontology

behaviour.dynamics_runtime.DynamicsRuntime.plot_ontology(**kwargs)

Plot this model’s ontology graph.

Parameters

Name	Type	Description	Default
**kwargs		Forwarded to tvbo.plot.ontology.plot_model.	{}

Returns

Name	Type	Description
		The rendered ontology plot.

render

behaviour.dynamics_runtime.DynamicsRuntime.render(format='yaml', **kwargs)

Unified entry point for rendering the model in any output format.

Dispatches to the appropriate renderer based on format:

• 'yaml' — TVBO YAML specification
• 'pyrates-yaml' — PyRates YAML
• 'report' / 'markdown' / 'md' — human-readable Markdown report
• 'pdf' — report rendered to PDF (requires outputfile kwarg)
• 'neuroml' / 'nml' / 'lems' — LEMS XML via NeuroMLAdapter
• Any code format accepted by :meth:render_code ('tvb', 'jax', 'julia', 'bifurcation-julia', …)

Parameters

format : str Target output format. **kwargs Forwarded to the underlying renderer.

Returns:

str

render_code

behaviour.dynamics_runtime.DynamicsRuntime.render_code(
    format='tvb',
    alt_label=None,
    **kwargs,
)

Generate backend source code for this model.

Dispatches to the template (or adapter) for the requested backend and returns the formatted source. Reads the model and does not modify it, so the source depends on the model alone and not on how often it has been rendered — nor on whether anything reordered it first. The dependency order the straight-line emitters need is a view the symbolic layer computes, not a state the model is put into: see in_dependency_order.

Parameters

Name	Type	Description	Default
format		Target backend, e.g. "tvb", "jax", "numpy", "tvboptim", "julia", "bifurcation-julia", "pde-fem", or "neuroml". The template-rendered ones are declared in CODE_FORMATS; the rest are built by an adapter.	'tvb'
alt_label		Optional alternative label for the generated model.	None
**kwargs		Forwarded to the template/adapter (e.g. continuation).	{}

Returns

Name	Type	Description
		The generated code as a formatted string.

Raises

Name	Type	Description
	ValueError	If format is not a supported backend.

render_equation

behaviour.dynamics_runtime.DynamicsRuntime.render_equation(
    obj,
    format='latex',
    inline_functions=False,
    **kwargs,
)

Render a model element’s equation to a string.

Handles conditional derived variables (converting conditionals to a SymPy Piecewise) and can optionally inline the model’s function definitions.

Parameters

Name	Type	Description	Default
obj		A model element exposing an equation (state/derived variable, derived parameter, …).	required
format		Output format, e.g. "latex", "numpy", "julia".	'latex'
inline_functions		If True, substitute model function bodies inline instead of emitting function calls.	False
**kwargs		Forwarded to tvbo.codegen.code.render_equation.	{}

Returns

Name	Type	Description
		The rendered equation in the requested format.

render_equation_cse

behaviour.dynamics_runtime.DynamicsRuntime.render_equation_cse(
    obj,
    format='numpy',
    inline_functions=False,
    **kwargs,
)

Common-subexpression-eliminated variant of :meth:render_equation.

Returns (setup, final) — a list of (name, expr) assignments plus the return expression — so interpreted backends (TVB / numpy) evaluate each shared subexpression (notably repeated model-function calls) once instead of per occurrence. Builds the same symbolic scope / user-function set as :meth:render_equation; see :func:tvbo.codegen.code.render_equation_cse.

run

behaviour.dynamics_runtime.DynamicsRuntime.run(
    format='python',
    verbose=0,
    save=True,
    run_kwargs=None,
    **kwargs,
)

Generate, execute, and integrate the model, returning its output.

Supports Julia (ODE and bifurcation), Python (SciPy odeint, or an iterated map for discrete systems), and compiled C backends.

Parameters

Name	Type	Description	Default
format		Backend to run, e.g. "python", "julia", "bifurcation-julia", or "c".	'python'
verbose		Verbosity level.	0
save		If True, cache results under a deterministic run filename.	True
run_kwargs		Extra arguments forwarded to the integrated dfun (e.g. stimulus).	None
**kwargs		Simulation settings such as duration, dt, t, and u_0.	{}

Returns

Name	Type	Description
	data_types.TimeSeries | analysis.BifurcationResult	A TimeSeries for time-domain
	data_types.TimeSeries | analysis.BifurcationResult	runs, or a BifurcationResult for bifurcation formats.

Raises

Name	Type	Description
	ValueError	If format is not supported.

save_model_metadata

behaviour.dynamics_runtime.DynamicsRuntime.save_model_metadata(filename)

Serialize the model metadata to a YAML file.

Parameters

Name	Type	Description	Default
filename		Destination path for the dumped YAML.	required

save_python_class

behaviour.dynamics_runtime.DynamicsRuntime.save_python_class(directory='.')

Write the model as a standalone TVB Python class file.

Emits <name>.py in directory with the required imports followed by the rendered TVB model code.

Parameters

Name	Type	Description	Default
directory		Target directory for the generated <name>.py file.	'.'

save_report

behaviour.dynamics_runtime.DynamicsRuntime.save_report(opath, format='markdown')

Generate the model report and write it to a directory.

Parameters

Name	Type	Description	Default
opath		Directory the report file is written to (as <name>.<ext>).	required
format		Report format passed to generate_report — "markdown" (written as .md) or "pdf".	'markdown'

search_ontology

behaviour.dynamics_runtime.DynamicsRuntime.search_ontology(search_str, **kwargs)

Search this model’s ontology subtree for a term.

Parameters

Name	Type	Description	Default
search_str	str	Text to search for among the model’s ontology labels and synonyms.	required
**kwargs		Forwarded to the underlying ontology search.	{}

Returns

Name	Type	Description
		The ontology search matches for search_str within this model.

symbolic_rhs

behaviour.dynamics_runtime.DynamicsRuntime.symbolic_rhs(obj, evaluate=True)

The parsed right-hand side of one of this model’s elements.

Resolved through the symbolic layer, so rendering an element reuses the expression already parsed for it rather than parsing its metadata again — the reason a second render_code on the same model costs nothing. Falls back to the element’s own Equation for anything the model does not declare (a stimulus, a caller’s ad-hoc element); parse_eq accepts either, so the caller does not need to know which.

evaluate must match what the caller would have parsed with. A backend that preserves authored term order needs the unevaluated form: SymPy canonicalises a + V*b + c*V**2 out of the order its author wrote it in, and the emitted source is compared against a frozen reference.

to_lems

behaviour.dynamics_runtime.DynamicsRuntime.to_lems(
    initial_conditions=1,
    component_id=None,
)

Build a LEMS model for this local neural mass model.

.. deprecated:: Use NeuroMLAdapter(model).render_code() from tvbo.adapters.neuroml instead. This method returns a lems.Model object (PyLEMS API); the adapter produces a validated XML string.

Parameters: - initial_conditions: number or dict; if number, used for all SVs; if dict, keys are sv name or sv_name_0 - component_id: optional id for the component; defaults to model label

Returns: - lems.Model instance containing a ComponentType and a Component for this model

to_pydantic

behaviour.dynamics_runtime.DynamicsRuntime.to_pydantic()

Return a tvbopydantic.Dynamics validated instance for this model.

to_yaml

behaviour.dynamics_runtime.DynamicsRuntime.to_yaml(filepath=None, format='tvbo')

Export the model to YAML format.

Parameters

filepath : str, optional Path to write the YAML file. If None, returns the YAML string. format : str Output format: “tvbo” (default) or “pyrates”. PyRates format generates a complete experiment YAML (model + network).

Returns:

str YAML string or filepath if written to file.

Example:

model.to_yaml(“model.yaml”) # TVBO format model.to_yaml(“model.yaml”, format=“pyrates”) # PyRates experiment format

update_parameters_from_equations

behaviour.dynamics_runtime.DynamicsRuntime.update_parameters_from_equations(
    default_value=1.0,
    overwrite=False,
)

Scan all equations and add any free symbols as parameters (default value if missing).

• Skips symbols that are known state variables, derived variables, or function arguments
• Skips the time symbol ‘t’
• Removes any previously added parameters that later become known entities
• Returns the list of parameter names that were added (or updated if overwrite=True)