Add RMSLE metric to extended_forecast_accuracy_metric_set function in yardstick-metric-sets.R script

[rserran]

Hi!

I've noticed that in Kaggle competitions, the RMSLE metric is used frequently instead of the RMSE.

This request is to add the RMSLE metric to the existing extended_forecast_accuracy_metric_set function in the yardstick-metric-sets.R script.

Julia Silge has a specific blog on the code to create this particular metric (see https://juliasilge.com/blog/nyc-airbnb/)

Code (very similar to maape code):

`library(rlang)

rmsle_vec <- function(truth, estimate, na_rm = TRUE, ...) {
rmsle_impl <- function(truth, estimate) {
sqrt(mean((log(truth + 1) - log(estimate + 1))^2))
}

metric_vec_template(
metric_impl = rmsle_impl,
truth = truth,
estimate = estimate,
na_rm = na_rm,
cls = "numeric",
...
)
}

rmsle <- function(data, ...) {
UseMethod("rmsle")
}
rmsle <- new_numeric_metric(rmsle, direction = "minimize")

rmsle.data.frame <- function(data, truth, estimate, na_rm = TRUE, ...) {
metric_summarizer(
metric_nm = "rmsle",
metric_fn = rmsle_vec,
data = data,
truth = !!enquo(truth),
estimate = !!enquo(estimate),
na_rm = na_rm,
...
)
}`

I can volunteer to add this code to the script.

Thanks for the attention and consideration to the request.

[mdancho84]

Hi yes, would you mind submitting a PR and we can discuss.