function
knowledge.function
Function Classes
Extended Function and LossFunction classes with code generation methods. Inherits from the LinkML datamodel and adds rendering/execution capabilities.
Usage
From YAML file:
from tvbo import Function, LossFunction
func = Function.from_file("correlation.yaml")
code = func.render_code(format='jax')
callable_fn = func.to_callable()From YAML string:
func = Function.from_string(yaml_string)From datamodel object:
from tvbo.datamodel import tvbo_datamodel
dm_func = tvbo_datamodel.Function(name='sigmoid', ...)
func = Function.from_datamodel(dm_func)Classes
| Name | Description |
|---|---|
| Function | Extended Function class with code generation and execution methods. |
| LossFunction | Extended LossFunction class with code generation and execution methods. |
Function
knowledge.function.Function(name='Function', **kwargs)Extended Function class with code generation and execution methods.
Inherits all schema fields from tvbo_datamodel.Function and adds: - Factory constructors: from_file, from_string, from_datamodel - Code generation: render_code, to_jax, to_numpy - Execution: to_callable
Attributes
| Name | Description |
|---|---|
| latex | Return LaTeX representation of the function equation. |
| sympy_expression | Return the parsed SymPy expression for this function’s equation. |
Methods
| Name | Description |
|---|---|
| from_datamodel | Create Function from a tvbo_datamodel.Function instance. |
| from_file | Load Function from a YAML file. |
| from_string | Load Function from a YAML string. |
| render_code | Generate Python code for this function. |
| to_callable | Generate and execute function code, returning the callable. |
| to_jax | Generate JAX code for this function. |
| to_numpy | Generate NumPy code for this function. |
| to_python | Generate pure Python code for this function. |
from_datamodel
knowledge.function.Function.from_datamodel(func)Create Function from a tvbo_datamodel.Function instance.
from_file
knowledge.function.Function.from_file(path)Load Function from a YAML file.
from_string
knowledge.function.Function.from_string(yaml_str)Load Function from a YAML string.
render_code
knowledge.function.Function.render_code(
format='jax',
user_functions=None,
render_func=None,
)Generate Python code for this function.
Parameters
format : str Output format: ‘jax’, ‘numpy’, ‘python’ user_functions : dict, optional Custom function name mappings for the printer. Example: {‘sigmoid’: ‘sigmoid’} to preserve function name render_func : callable, optional Custom render function for model context.
Returns
str Python code string defining the function
Examples
func = Function.from_string(yaml_str) print(func.render_code()) def sigmoid(x): return 1/(1 + jnp.exp(-x))
to_callable
knowledge.function.Function.to_callable(
format='jax',
user_functions=None,
namespace=None,
)Generate and execute function code, returning the callable.
Parameters
format : str Output format: ‘jax’, ‘numpy’ user_functions : dict, optional Custom function name mappings namespace : dict, optional Namespace for exec(). If None, creates one with jnp/np imports.
Returns
callable The generated function as a callable
Examples
func = Function.from_string(sigmoid_yaml) sigmoid = func.to_callable() sigmoid(0.0) 0.5
to_jax
knowledge.function.Function.to_jax(**kwargs)Generate JAX code for this function.
to_numpy
knowledge.function.Function.to_numpy(**kwargs)Generate NumPy code for this function.
to_python
knowledge.function.Function.to_python(**kwargs)Generate pure Python code for this function.
LossFunction
knowledge.function.LossFunction(name='LossFunction', **kwargs)Extended LossFunction class with code generation and execution methods.
Inherits all schema fields from tvbo_datamodel.LossFunction and adds: - Factory constructors: from_file, from_string, from_datamodel - Code generation: render_code, to_jax, to_numpy - Execution: to_callable
LossFunction extends Function with aggregation specification for per-element losses (e.g., mean over nodes).
Attributes
| Name | Description |
|---|---|
| aggregation_dimension | Return the aggregation dimension as a string (e.g., ‘node’). |
| aggregation_type | Return the aggregation type as a string (e.g., ‘mean’, ‘sum’). |
| latex | Return LaTeX representation of the loss function equation. |
| sympy_expression | Return the parsed SymPy expression for this function’s equation. |
Methods
| Name | Description |
|---|---|
| from_datamodel | Create LossFunction from a tvbo_datamodel.LossFunction instance. |
| from_file | Load LossFunction from a YAML file. |
| from_string | Load LossFunction from a YAML string. |
| render_code | Generate Python code for this loss function with aggregation. |
| to_callable | Generate and execute loss function code, returning the callable. |
| to_jax | Generate JAX code for this loss function. |
| to_numpy | Generate NumPy code for this loss function. |
from_datamodel
knowledge.function.LossFunction.from_datamodel(func)Create LossFunction from a tvbo_datamodel.LossFunction instance.
from_file
knowledge.function.LossFunction.from_file(path)Load LossFunction from a YAML file.
from_string
knowledge.function.LossFunction.from_string(yaml_str)Load LossFunction from a YAML string.
render_code
knowledge.function.LossFunction.render_code(
format='jax',
user_functions=None,
inner_func_names=None,
)Generate Python code for this loss function with aggregation.
Parameters
format : str Output format: ‘jax’, ‘numpy’ user_functions : dict, optional Custom function name mappings inner_func_names : list, optional Names of inner functions that should be recognized. Example: [‘correlation’] for “1 - correlation(x, y)”
Returns
str Python code string defining the loss function
Examples
loss = LossFunction.from_string(loss_yaml) print(loss.render_code(inner_func_names=[‘correlation’])) def spectral_loss(sim, target): def _per_element_loss(sim, target): return 1 - correlation(sim, target) per_element_losses = jax.vmap(_per_element_loss)(sim, target) return jnp.mean(per_element_losses)
to_callable
knowledge.function.LossFunction.to_callable(
format='jax',
user_functions=None,
inner_func_names=None,
namespace=None,
)Generate and execute loss function code, returning the callable.
Parameters
format : str Output format: ‘jax’, ‘numpy’ user_functions : dict, optional Custom function name mappings inner_func_names : list, optional Names of inner functions to recognize namespace : dict, optional Namespace for exec(). If None, creates one with jnp/np/jax imports.
Returns
callable The generated loss function as a callable
to_jax
knowledge.function.LossFunction.to_jax(**kwargs)Generate JAX code for this loss function.
to_numpy
knowledge.function.LossFunction.to_numpy(**kwargs)Generate NumPy code for this loss function.