parameter.LogPositiveParameter

parameter.LogPositiveParameter(value, lower=0.0)

Parameter constrained to be positive using log transformation.

Maps from unconstrained real space to (lower, ∞) using log transformation. This provides smooth gradients and is ideal for scale parameters like conductances, time constants, etc.

Parameters

Name	Type	Description	Default
value	Union[float, int, jnp.ndarray]	The initial parameter value (must be > lower).	required
lower	float	Lower bound for the parameter value. Default is 0.0.	`0.0`

Examples

>>> param = LogPositiveParameter(2.0)  # Always positive
>>> param = LogPositiveParameter(5.0, lower=1.0)  # Always > 1.0

Methods

Name	Description
tree_flatten	Flatten for JAX pytree registration.
tree_unflatten	Unflatten for JAX pytree registration.

tree_flatten

parameter.LogPositiveParameter.tree_flatten()

Flatten for JAX pytree registration.

tree_unflatten

parameter.LogPositiveParameter.tree_unflatten(aux_data, children)

Unflatten for JAX pytree registration.

# parameter.LogPositiveParameter { #tvboptim.parameter.LogPositiveParameter }

```python
parameter.LogPositiveParameter(value, lower=0.0)
```

Parameter constrained to be positive using log transformation.

Maps from unconstrained real space to (lower, ∞) using log transformation.
This provides smooth gradients and is ideal for scale parameters like
conductances, time constants, etc.

## Parameters {.doc-section .doc-section-parameters}

| Name   | Type                                                                                                         | Description                                          | Default    |
|--------|--------------------------------------------------------------------------------------------------------------|------------------------------------------------------|------------|
| value  | [Union](`typing.Union`)\[[float](`float`), [int](`int`), [jnp](`jax.numpy`).[ndarray](`jax.numpy.ndarray`)\] | The initial parameter value (must be > lower).       | _required_ |
| lower  | [float](`float`)                                                                                             | Lower bound for the parameter value. Default is 0.0. | `0.0`      |

## Examples {.doc-section .doc-section-examples}

```python
>>> param = LogPositiveParameter(2.0)  # Always positive
>>> param = LogPositiveParameter(5.0, lower=1.0)  # Always > 1.0
```

## Methods

| Name | Description |
| --- | --- |
| [tree_flatten](#tvboptim.parameter.LogPositiveParameter.tree_flatten) | Flatten for JAX pytree registration. |
| [tree_unflatten](#tvboptim.parameter.LogPositiveParameter.tree_unflatten) | Unflatten for JAX pytree registration. |

### tree_flatten { #tvboptim.parameter.LogPositiveParameter.tree_flatten }

```python
parameter.LogPositiveParameter.tree_flatten()
```

Flatten for JAX pytree registration.

### tree_unflatten { #tvboptim.parameter.LogPositiveParameter.tree_unflatten }

```python
parameter.LogPositiveParameter.tree_unflatten(aux_data, children)
```

Unflatten for JAX pytree registration.