scoringutils (version 2.1.0)

interval_score: Interval score

Description

Proper Scoring Rule to score quantile predictions, following Gneiting and Raftery (2007). Smaller values are better.

The score is computed as

$$ \textrm{score} = (\textrm{upper} - \textrm{lower}) + \frac{2}{\alpha}(\textrm{lower} - \textrm{observed}) * \mathbf{1}(\textrm{observed} < \textrm{lower}) + \frac{2}{\alpha}(\textrm{observed} - \textrm{upper}) * \mathbf{1}(\textrm{observed} > \textrm{upper}) $$ where \(\mathbf{1}()\) is the indicator function and indicates how much is outside the prediction interval. \(\alpha\) is the decimal value that indicates how much is outside the prediction interval.

To improve usability, the user is asked to provide an interval range in percentage terms, i.e. interval_range = 90 (percent) for a 90 percent prediction interval. Correspondingly, the user would have to provide the 5% and 95% quantiles (the corresponding alpha would then be 0.1). No specific distribution is assumed, but the interval has to be symmetric around the median (i.e you can't use the 0.1 quantile as the lower bound and the 0.7 quantile as the upper bound). Non-symmetric quantiles can be scored using the function quantile_score().

Usage

interval_score(
  observed,
  lower,
  upper,
  interval_range,
  weigh = TRUE,
  separate_results = FALSE
)

Value

Vector with the scoring values, or a list with separate entries if separate_results is TRUE.

Arguments

observed

A vector with observed values of size n

lower

Vector of size n with the prediction for the lower quantile of the given interval range.

upper

Vector of size n with the prediction for the upper quantile of the given interval range.

interval_range

Numeric vector (either a single number or a vector of size n) with the range of the prediction intervals. For example, if you're forecasting the 0.05 and 0.95 quantile, the interval range would be 90. The interval range corresponds to \((100-\alpha)/100\), where \(\alpha\) is the decimal value that indicates how much is outside the prediction interval (see e.g. Gneiting and Raftery (2007)).

weigh

Logical. If TRUE (the default), weigh the score by \(\alpha / 2\), so it can be averaged into an interval score that, in the limit (for an increasing number of equally spaced quantiles/prediction intervals), corresponds to the CRPS. \(\alpha\) is the value that corresponds to the (\(\alpha/2\)) or (\(1 - \alpha/2\)), i.e. it is the decimal value that represents how much is outside a central prediction interval (E.g. for a 90 percent central prediction interval, alpha is 0.1).

separate_results

Logical. If TRUE (default is FALSE), then the separate parts of the interval score (dispersion penalty, penalties for over- and under-prediction get returned as separate elements of a list). If you want a data.frame instead, simply call as.data.frame() on the output.

References

Strictly Proper Scoring Rules, Prediction,and Estimation, Tilmann Gneiting and Adrian E. Raftery, 2007, Journal of the American Statistical Association, Volume 102, 2007 - Issue 477

Evaluating epidemic forecasts in an interval format, Johannes Bracher, Evan L. Ray, Tilmann Gneiting and Nicholas G. Reich, https://journals.plos.org/ploscompbiol/article?id=10.1371/journal.pcbi.1008618 # nolint

Examples

observed <- rnorm(30, mean = 1:30)
interval_range <- rep(90, 30)
alpha <- (100 - interval_range) / 100
lower <- qnorm(alpha / 2, rnorm(30, mean = 1:30))
upper <- qnorm((1 - alpha / 2), rnorm(30, mean = 11:40))

scoringutils:::interval_score(
  observed = observed,
  lower = lower,
  upper = upper,
  interval_range = interval_range
)

# gives a warning, as the interval_range should likely be 50 instead of 0.5
scoringutils:::interval_score(
  observed = 4, upper = 8, lower = 2, interval_range = 0.5
)

# example with missing values and separate results
scoringutils:::interval_score(
  observed = c(observed, NA),
  lower = c(lower, NA),
  upper = c(NA, upper),
  separate_results = TRUE,
  interval_range = 90
)