https://github.com/devmotion/rcalibration
Estimation and hypothesis tests of calibration in R using CalibrationErrors.jl and CalibrationTests.jl.
https://github.com/devmotion/rcalibration
calibration julia-language r-language reliability
Last synced: 5 months ago
JSON representation
Estimation and hypothesis tests of calibration in R using CalibrationErrors.jl and CalibrationTests.jl.
- Host: GitHub
- URL: https://github.com/devmotion/rcalibration
- Owner: devmotion
- License: other
- Created: 2021-05-03T23:46:23.000Z (about 5 years ago)
- Default Branch: main
- Last Pushed: 2022-02-12T22:40:33.000Z (over 4 years ago)
- Last Synced: 2025-06-16T13:12:19.551Z (about 1 year ago)
- Topics: calibration, julia-language, r-language, reliability
- Language: R
- Homepage: https://devmotion.github.io/CalibrationErrors.jl/dev
- Size: 28.3 KB
- Stars: 1
- Watchers: 2
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# rcalibration
Estimation and hypothesis tests of calibration in R using CalibrationErrors.jl and CalibrationTests.jl.
[](https://devmotion.github.io/CalibrationErrors.jl/stable)
[](https://devmotion.github.io/CalibrationErrors.jl/dev)
[](https://github.com/devmotion/rcalibration/actions?query=workflow%3AR-CMD-check+branch%3Amain)
[](https://github.com/devmotion/CalibrationErrors.jl/actions?query=workflow%3ACI+branch%3Amain)
[](https://github.com/devmotion/CalibrationTests.jl/actions?query=workflow%3ACI+branch%3Amain)
rcalibration is a package for estimating calibration of probabilistic models in R.
It is an R interface for [CalibrationErrors.jl](https://github.com/devmotion/CalibrationErrors.jl) and [CalibrationTests.jl](https://github.com/devmotion/CalibrationTests.jl).
As such, the package allows the estimation of calibration errors (ECE and SKCE) and statistical testing of the null hypothesis that a model is calibrated.
## Installation
You can install `rcalibration` with [`devtools`](https://devtools.r-lib.org/):
```R
> library(devtools)
> devtools::install_github("devmotion/rcalibration")
```
The use of `rcalibration` requires that its dependency
[`JuliaCall`](https://github.com/Non-Contradiction/JuliaCall) (installed automatically)
and itself are configured correctly.
For `JuliaCall`, you have to
[install Julia](https://github.com/Non-Contradiction/JuliaCall#installation).
The configuration process is described in the
[`JuliaCall` documentation](https://non-contradiction.github.io/JuliaCall/index.html).
When `JuliaCall` is configured correctly, you can install the Julia packages required by
`rcalibration`:
```R
> library(rcalibration)
> rcalibration::install()
```
### Crash on MacOS with Julia 1.6
Due to a [problem in Julia 1.6](https://github.com/JuliaLang/julia/issues/40246), `JuliaCall`
and `rcalibration` crash on MacOS with this Julia version. Please use Julia 1.5 on MacOS
until this issue is fixed.
### Custom Julia environment
With the default settings, `JuliaCall` and `rcalibration` install all Julia dependencies
in the default environment. In particular, if you use Julia for other projects as well,
a separate [project environment](https://pkgdocs.julialang.org/v1/environments/) can
simplify package management and ensure that the state of the Julia dependencies is
reproducible. In `JuliaCall` and `rcalibration`, a custom project environment is used if
you set the environment variable `JULIA_PROJECT`:
```shell
export JULIA_PROJECT="path/to/the/environment/"
```
## Usage
Import and setup calibration analysis tools from CalibrationErrors.jl and CalibrationTests.jl with
```R
> ca <- rcalibration::load()
```
You can then do the same as would be done in Julia, except you have to add `ca$` in front for functionality from the Julia packages.
Most of the commands will work without any modification.
Thus the documentation of the Julia packages is the main in-depth documentation for this package.
### Callable objects
R does not support the callable object syntax that is a common idiom in Julia.
`JuliaCall` supports the
[syntax `f$.(x)` in R for the function call `f(x)`](https://github.com/Non-Contradiction/JuliaCall/pull/118#issuecomment-534203455)
with callable object `f` in Julia.
### Calibration errors
Let us estimate the squared kernel calibration error (SKCE) with the tensor
product kernel
```math
k((p, y), (p̃, ỹ)) = exp(-|p - p̃|) δ(y - ỹ)
```
from a set of predictions and corresponding observed outcomes.
```R
> skce <- ca$SKCE(ca$tensor(ca$ExponentialKernel(), ca$WhiteKernel()))
```
Other estimators of the SKCE and estimators of other calibration errors such
as the expected calibration error (ECE) are available as well. The Julia package
[KernelFunctions.jl](https://github.com/JuliaGaussianProcesses/KernelFunctions.jl)
supports a variety of kernels, all compositions and transformations of
[kernels available there](https://juliagaussianprocesses.github.io/KernelFunctions.jl/stable/kernels/)
can be used.
#### Probabilities
Predictions can be provided as probabilities. In this case, the
predictions correspond to Bernoulli distributions with these parameters and the
targets are boolean values.
```R
> set.seed(1234)
> predictions <- runif(100)
> outcomes <- sample(c(TRUE, FALSE), 100, replace=TRUE)
> skce$.(predictions, outcomes)
[1] 0.01518318
```
#### Probability vectors
Predictions can be provided as probability vectors (i.e., vectors in the probability
simplex) as well. In this case, the predictions correspond to categorical
distributions with these class probabilities and the targets are integers in `{1,...,n}`.
The probability vectors can be given as a matrix. However, it is
required to specify if the probability vectors correspond to rows or columns of the matrix
by wrapping them in `ca.RowVecs` and `ca.ColVecs`, respectively. These wrappers are defined
in [KernelFunctions.jl](https://github.com/JuliaGaussianProcesses/KernelFunctions.jl).
```R
> library(extraDistr)
> set.seed(1234)
> predictions <- rdirichlet(100, c(3, 2, 5))
> outcomes <- sample(1:3, 100, replace=TRUE)
> skce$.(ca$RowVecs(predictions), outcomes)
[1] 0.02585344
```
#### Probability distributions
Predictions can also be provided as probability distributions defined in the
Julia package [Distributions.jl](https://github.com/JuliaStats/Distributions.jl). Currently,
analytical formulas for the estimators of the SKCE and unnormalized calibration mean embedding
(UCME) are implemented for uni- and multivariate normal distributions `ca$Normal` and
`ca$MvNormal` with squared exponential kernels on the target space and Laplace distributions
`ca$Laplace` with exponential kernels on the target space.
In this example we use the tensor product kernel
```math
k((p, y), (p̃, ỹ)) = exp(-W₂(p, p̃)) exp(-(y - ỹ)²/2),
```
where `W₂(p, p̃)` is the 2-Wasserstein distance of the two normal distributions `p` and `p̃`.
It is given by
```math
W₂(p, p̃) = √((μ - μ̃)² + (σ - σ̃)²),
```
where `p = N(μ, σ)` and `p̃ = N(μ̃, σ̃)`.
```R
> set.seed(1234)
> predictions <- replicate(100, ca$Normal(rnorm(1), runif(1)))
> outcomes <- rnorm(100)
> skce <- ca$SKCE(ca$tensor(ca$ExponentialKernel(metric=ca$Wasserstein()), ca$SqExponentialKernel()))
> skce$.(predictions, outcomes)
[1] 0.02301165
```
### Calibration tests
`rcalibration` provides different calibration tests that estimate the p-value of the null hypothesis
that a model is calibrated, based on a set of predictions and outcomes:
- `ca$ConsistencyTest` estimates the p-value with consistency resampling for a given calibration error estimator
- `ca$DistributionFreeSKCETest` computes distribution-free (and therefore usually quite weak) upper bounds of the p-value for different estimators of the SKCE
- `ca$AsymptoticBlockSKCETest` estimates the p-value based on the asymptotic distribution of the unbiased block estimator of the SKCE
- `ca$AsymptoticSKCETest` estimates the p-value based on the asymptotic distribution of the unbiased estimator of the SKCE
- `ca$AsymptoticCMETest` estimates the p-value based on the asymptotic distribution of the UCME
```R
> library(extraDistr)
> set.seed(1234)
> predictions <- rdirichlet(100, c(3, 2, 5))
> outcomes <- sample(1:3, 100, replace=TRUE)
> test <- ca$AsymptoticSKCETest(kernel, ca$RowVecs(predictions), outcomes)
> print(test)
Julia Object of type AsymptoticSKCETest{KernelTensorProduct{Tuple{ExponentialKernel{TotalVariation}, WhiteKernel}}, Float64, Float64, Matrix{Float64}}.
Asymptotic SKCE test
--------------------
Population details:
parameter of interest: SKCE
value under h_0: 0.0
point estimate: 0.0259434
Test summary:
outcome with 95% confidence: reject h_0
one-sided p-value: 0.0100
Details:
test statistic: -0.007291403994633658
> ca$pvalue(test)
[1] 0.004
```
## Citing
If you use rcalibration as part of your research, teaching, or other activities, please consider citing the following publications:
Widmann, D., Lindsten, F., & Zachariah, D. (2019). [Calibration tests in multi-class classification: A unifying framework](https://proceedings.neurips.cc/paper/2019/hash/1c336b8080f82bcc2cd2499b4c57261d-Abstract.html). In *Advances in Neural Information Processing Systems 32 (NeurIPS 2019)* (pp. 12257–12267).
Widmann, D., Lindsten, F., & Zachariah, D. (2021). [Calibration tests beyond classification](https://openreview.net/forum?id=-bxf89v3Nx). *International Conference on Learning Representations (ICLR 2021)*.
## Acknowledgements
This work was financially supported by the Swedish Research Council via the projects *Learning of Large-Scale Probabilistic Dynamical Models* (contract number: 2016-04278), *Counterfactual Prediction Methods for Heterogeneous Populations* (contract number: 2018-05040), and *Handling Uncertainty in Machine Learning Systems* (contract number: 2020-04122), by the Swedish Foundation for Strategic Research via the project *Probabilistic Modeling and Inference for Machine Learning* (contract number: ICA16-0015), by the Wallenberg AI, Autonomous Systems and Software Program (WASP) funded by the Knut and Alice Wallenberg Foundation, and by ELLIIT.