Relativeaveragespectralerror

---
name: relativeaveragespectralerror
description: Compute the RelativeAverageSpectralError metric — provided by torchmetrics. Use when the user has predictions and ground-truth and needs to compute RelativeAverageSpectralError, or asks how to score with RelativeAverageSpectralError.
metadata:
  skill_kind: metric
  source_lib: torchmetrics
  import_path: torchmetrics.RelativeAverageSpectralError
  source: library_introspection
---

# relativeaveragespectralerror

> Metric `RelativeAverageSpectralError` from `torchmetrics` (torchmetrics.RelativeAverageSpectralError)

## When to invoke this skill

The user has predictions + ground truth and asks to evaluate with RelativeAverageSpectralError, or
mentions `torchmetrics.RelativeAverageSpectralError` directly, or wants the standard torchmetrics implementation.

## Reference signature

```python
from torchmetrics import RelativeAverageSpectralError

# _RelativeAverageSpectralError(window_size: int = 8, **kwargs: dict[str, typing.Any]) -> None
```

## Library docstring

```
Wrapper for deprecated import.

>>> from torch import rand
>>> preds = rand(4, 3, 16, 16)
>>> target = rand(4, 3, 16, 16)
>>> rase = _RelativeAverageSpectralError()
>>> rase(preds, target)
tensor(5326.40...)
```

## Quick recipe

```python
import torchmetrics as _m
score = _m.RelativeAverageSpectralError(y_true, y_pred)
```

## Don'ts

- Don't reimplement when the library version handles edge cases (NaN, ties, empty inputs) better than a hand-rolled formula.
- Always check the library version's argument order — sklearn is `(y_true, y_pred)` while torchmetrics is `(preds, target)`.