scoringutils

scoringutils copied to clipboard

Use nice illustration of CRPS/DSS/Log score + the number of samples needed for scores to converge

1

We have this visualisation, which among others makes the point that you need quite a lot of samples for the sample-based log score to be reliable.  It was previously...

nikosbosse

Use illustration of local and global scores

1

We have this nice illustration:  It's not used in the manuscript anymore. But it would be nice to use it in a vignette explaining differences between scores. Related: #933,...

nikosbosse

Use nice illustration of Sharpness and calibration

1

We have this illustration:  Generated by this code: ``` library(ggplot2) library(patchwork) library(scoringutils) p1

nikosbosse

Make all internal columns, in particular `forecast_id`, more robust by prefixing it with "scoringutils_"

1

`quantile_to_interval_numeric()`, for example, makes use of an internal `forecast_id` column. This might cause issues if there is already a `forecast_id` column in the input. We should likely prefix all internal...

nikosbosse

(but happy to leave as is and we should maybe just open an issue to get rid of `check_columns_present` altogether) _Originally posted by @nikosbosse in https://github.com/epiforecasts/scoringutils/pull/915#discussion_r1768078954_ Seems like functionality can...

sbfnk

coverage of function factory

5

As mentioned in https://github.com/hubverse-org/hubEvals/pull/46 do we want to have any kind of special metric function factories for generating things like multiple interval coverages? I think the answer is potentially with...

seabbs

Compact version of `get_duplicate_forecasts()`

1

Maybe something like the following would be nice. Normal output ``` ex

nikosbosse