The goal of tfNeuralODE is to provide an easy framework surrounding the use of Neural Ordinary Differential Equations in R. This package does so by building on top of R Tensorflow, to help wth auto-differentiation and the construction of neural network models with Keras. Examples of this package in use can be found on the project website.
Please note that in its current state, tfNeuralODE is in not an optimized state, with programs running quite slowly at the moment. Improvements to speed are a priority on the development agenda.
You can install the development version of tfNeuralODE like so:
devtools::install_github("https://github.com/semran9/tfNeuralODE.git")A release version can be installed from CRAN.
install.packages("tfNeuralODE")