noramvillanueva / keras-workshop

Data and scripts for keras course

Keras for R Workshop

To follow along with the workshop you have two options:

1. We will provide a cloud-base RStudio environment with everything pre-installed and datasets in place. We'll give you a URL and a login on the day and you'll only need a laptop with a web browser and ability to get on wifi.
2. All of the models we're building will work on a laptop so if you want to follow along on your own machine then please follow the steps below:

• Install R from https://cran.r-project.org/
• RStudio desktop from https://www.rstudio.com

Install the following R packages from CRAN in the usual way:

install.packages(c("tidyverse", "rsample", "keras"))

You'll also need the devel version of recipes:

require("devtools")
install_github("topepo/recipes")

In an R session install the keras/tensorflow python libraries by running:

library(keras)
install_keras()

This can take a while as it will install anaconda python at the same time. For further instructions please see https://keras.rstudio.com/ and follow the instructions there.

If it worked you should get:

library(keras)
is_keras_available()
[1] TRUE