R 1: Loading the Data

First, we load the necessary software dependencies. This means you need to use reticulate to find your python path:

reticulate:::conda_list(conda = "auto")

Then using that python path to source the package for R to use it:

library(interactiontransformer)
source.python('/anaconda2/envs/py36/bin/python')
interactiontransformer<-import_transformer()

Then, let's access some test data using this command:

df<-read.csv("../test_data/epistasis.test.csv")
X<-df[,-(ncol(df))]
y<-as.data.frame(df[,ncol(df)])

Now that our data is loaded, we can split the data up into training and test sets, though if you are performing traditional statistical analyses, you do not need to conduct this split:

train.test.splits<-train.test.split(X,y)

Now, we will fit the interaction transformer!