cvms 1.0.0 released
After 6 months of work,
1.0.0 has finally been released!
This version is a major refactoring of the package and includes tons of new features and changes .
The most important additions are:
- Hyperparameter tuning of custom model functions
- Within-cv preprocessing
- Multiple new metrics
- Identification of observations that are difficult to predict
- Four new vignettes (tutorials)
cvms no longer depends on
caret, as the creation of confusion matrices and calculation of related metrics are now implemented in
cvms. This should make installation easier.
cross_validate_fn() has been improved and should allow cross-validation of most model functions.
There’s a list of breaking changes here .
- A big one is that the
validate()has been renamed to
formulasto be consistent with
- Another big one is that the
typeargument no longer has a default value.
cvms 1.0.0 📦 is now on CRAN! 😍— Ludvig Olsen (@LudvigOlsen) April 14, 2020
Major refactoring and tons of new stuff. A few breaking changes to be aware of.
news: https://t.co/tMsBiMyukX#rstats #statistics #DataScience #R #MachineLearning
One of my new favorite functions is
select_definitions() makes it faster to extract relevant columns from the output when comparing the models:
I should also mention select_definitions(), which makes it easier to read the cross-validation output. It selects the columns that define the model (formula parts) and unnests the applied hyperparameters. We can include a couple of metrics as well.#rstats— Ludvig Olsen (@LudvigOlsen) April 14, 2020