The hardware and bandwidth for this mirror is donated by METANET, the Webhosting and Full Service-Cloud Provider.
If you wish to report a bug, or if you are interested in having us mirror your free-software or open-source project, please feel free to contact us at mirror[@]metanet.ch.
A clinical prediction model should produce calibrated risk
predictions, which means the predicted probabilities should align with
observed probabilities. There are various ways of assessing calibration
(this paper
covers calibration in more detail). pmcalibration
implements calibration curves for binary and (right censored)
time-to-event outcomes and calculates metrics used to assess the
correspondence between predicted and observed outcome probabilities (the
‘integrated calibration index’ or \(ICI\), aka \(E_{avg}\), as well as \(E_{50}\), \(E_{90}\), and \(E_{max}\)).
A goal of pmcalibration
is to implement a range of
methods for estimating a smooth relationship between predicted and
observed probabilities and to provide confidence intervals for
calibration metrics (via bootstrapping or simulation based
inference).
To install:
install.packages("pmcalibration")
To install development version:
devtools::install_github("https://github.com/stephenrho/pmcalibration")
These binaries (installable software) and packages are in development.
They may not be fully stable and should be used with caution. We make no claims about them.