README.rmd

---
title: "README.rmd"
author: "Jake S. Rhodes"
date: "2/9/2022"
output: rmarkdown::github_document
---

```{r setup, include=FALSE}
knitr::opts_chunk$set(echo = TRUE)
```

#  RF-GAP
## Random Forest Geometry- and Accuracy-Preserving proximities

This is the official repository for the papr "Random Forest- Geometry- and Accuracy-Preserving Proximities" (https://arxiv.org/abs/2201.12682). In the paper we show that random forest (RF) predictions can be exactly determined by using RF-GAP proximities as weights in a weighted-sum regressor or weighted-majority vote classifier.  This repo provides the base code to generate the various proximity definitions described in the paper. We provide some examples below.


## Generate RF-GAP proximities:

```{r rfgap}
library(rfgap)

x <- iris[, 1:4]
y <- iris[, 5]
prox <- get_proximities(x, y, type = 'rfgap')
```

## Create 2-dimensional MDS embedding and plot

```{r}
x <- iris[, 1:4]
y <- iris[, 5]
mds <- rf_mds(x, y, type = 'rfgap')

plot(mds, y)
```


## Impute missing data
```{r}
x <- airquality[, -4]
y <- airquality[, 4]
imputed_data <- rf_impute(x, y, type = 'rfgap')
```

## Run Outlier Detection
```{r}
x <- mtcars[, -c(1, 2)]
y <- as.factor(mtcars[, 2])
outlier_scores <- rf_outliers(x, y, type = 'rfgap')

plot(outlier_scores, x, y)
```