A configurable evaluator for regression tasks. The function computes
various regression metrics, particularly useful for reconstruction tasks
(e.g., autoencoders). Reproduces the behaviour of Python's
regression_metrics_fn.
Supported metrics names:
"kl_divergence" – Kullback-Leibler divergence
"mse" – Mean Squared Error
"mae" – Mean Absolute Error
Usage
regression_metrics_fn(x, x_rec, metrics = NULL)
Arguments
- x
Numeric vector, matrix, or torch tensor of true target data sample.
- x_rec
Numeric vector, matrix, or torch tensor of reconstructed data sample.
- metrics
Character vector listing which metrics to compute. Default
is c("kl_divergence", "mse", "mae").
Value
Named numeric vector with one element per requested metric.
Examples
set.seed(42)
x <- runif(1000)
x_rec <- runif(1000)
regression_metrics_fn(x, x_rec, metrics = c("mse", "mae"))
#> mse mae
#> 0.1726048 0.3409409