## SelectiveMetric


Evaluate a metric on full, selected, and rejected subsets.


Usage

``` python
SelectiveMetric(base)
```


Wraps a `torchmetrics.Metric` or `torchmetrics.MetricCollection` and keeps three independent copies that are updated separately:

- `"full"`: all samples passed to [update](SelectiveMetric.md#seapig.SelectiveMetric.update).
- `"selected"`: samples where the provided selection mask is true.
- `"rejected"`: samples where the selection mask is false.

The [compute](SelectiveMetric.md#seapig.SelectiveMetric.compute) result is a flat `dict`\[`str`, `torch.Tensor`\] where each underlying metric name is prefixed with `full/`, `selected/`, or `rejected/`. If a submetric was never updated, its value is a zero `torch.Tensor`.


## Parameters


`base: torchmetrics.Metric | torchmetrics.MetricCollection`  
Metric (or collection) to wrap. Internally the object is deep- copied three times so each subset is tracked independently.


## Notes

- The selection mask may be boolean or numeric; numeric values `> 0` are treated as selected.
- Calls that contain no selected (or no rejected) rows do not update the corresponding internal metric for that call.


## Example


``` python
import torch
from torchmetrics import Accuracy
from seapig import SelectiveMetric

base = Accuracy(task="binary")
m = SelectiveMetric(base)
preds = torch.tensor([[0.9, 0.1], [0.2, 0.8]])
target = torch.tensor([[1.0, 1], [1, 0]])
selected = torch.tensor([1, 0], dtype=torch.bool)
m.update(preds, target, selected)
results = m.compute()
print(results)
```


    {'full/BinaryAccuracy': tensor(0.2500), 'selected/BinaryAccuracy': tensor(0.5000), 'rejected/BinaryAccuracy': tensor(0.)}


## Methods

| Name | Description |
|----|----|
| [compute()](#compute) | Compute metrics and return a dict with prefixed keys. |
| [reset()](#reset) | Reset the internal metric instances. |
| [update()](#update) | Update full, selected, and rejected metrics. |

------------------------------------------------------------------------


#### compute()


Compute metrics and return a dict with prefixed keys.


Usage

``` python
compute()
```


Returns a dictionary where each key is `<scope>/<metric_name>` (scope is one of `full`, `selected`, `rejected`). Values are `torch.Tensor`s. If a metric instance was never updated, its value is a scalar zero tensor.


------------------------------------------------------------------------


#### reset()


Reset the internal metric instances.


Usage

``` python
reset()
```


------------------------------------------------------------------------


#### update()


Update full, selected, and rejected metrics.


Usage

``` python
update(preds, target, selected)
```


##### Parameters


`preds: torch.Tensor`  
Model predictions of shape (B, …).

`target: torch.Tensor`  
Target tensor of shape (B, …).

`selected: torch.Tensor`  
Boolean or binary selection mask of shape (B,). Values \> 0 are treated as selected.