---
title: "keras - Neural Network"
output: html_document
vignette: >
  %\VignetteEncoding{UTF-8}
  %\VignetteIndexEntry{keras - Neural Network}
  %\VignetteEngine{knitr::rmarkdown}
editor_options:
  chunk_output_type: console
---


```{r, include = FALSE}
knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>",
  fig.path = "vig/"
)
options(rmarkdown.html_vignette.check_title = FALSE)
```

<!-- avoid border around images -->
<style>
img {
  border: 0;
  }
</style>

```{r, message=FALSE}
library('vivid')
library("keras")
```

This guide is designed as a quick-stop reference of how to use some of the more popular machine learning R packages with `vivid`. In the following example, we use the air quality data for regression.


### keras - Neural Network
The `keras` package in R is an interface to the original `keras` Python library. It allows R users to build and train neural network models using the high-level Keras API. To use `keras` the data must first be normalised. Due to the complexity custom model fits in `keras`, producing a custom predict function to use with `vivid` may require some work. In the example below, we show how to use `vivid` when building a `keras` model composed of a linear stack of layers (i.e., using ` keras_model_sequential`). Additionally, currently the PDP plots in `vivid` only work for the development version on GitHub.



```{r, eval=FALSE}
# load data
aq <- na.omit(airquality)
# Set up data
train_data <- aq %>%
  as.matrix()
train_targets <- aq$Ozone

# Normalize the data
mean <- apply(train_data, 2, mean)                                  
std <- apply(train_data, 2, sd)
aqTrain <- scale(train_data, center = mean, scale = std)  

# Define model architecture
build_model <- function() {                                
  model <- keras_model_sequential() %>%
    layer_dense(units = 64, activation = "relu",
                input_shape = dim(aqTrain)[[2]]) %>%
    layer_dense(units = 64, activation = "relu") %>%
    layer_dense(units = 1)
  model %>% compile(
    optimizer = "rmsprop",
    loss = "mse",
    metrics = c("mae")
  )
}

# Train the final model
model <- build_model()
model %>% fit(train_data, train_targets, epochs = 80, 
              batch_size = 16, verbose = 0)

# Create usable data frame from scaled data
aqFinal <- as.data.frame(aqTrain)

# predict function for keras
vi_fun <- function(fit, data,...) {
  predict(fit, x = as.matrix(data))
}

# vivid
vi <- vivi(data = aqFinal, 
           fit = model,
           response = 'Ozone',
           predictFun = vi_fun)

```


#### Heatmap
```{r, ke_r_heat,  out.width = '100%', eval=FALSE}
viviHeatmap(mat = vi)
```


```{r, echo = F,  out.width = '100%'}
knitr::include_graphics("https://raw.githubusercontent.com/AlanInglis/vivid/master/vignettes/vig/ke_r_heat-1.png")
```
<center>
<caption><span id="fig1:fig1">Figure 1: </span>Heatmap of a keras neural network regression fit displaying 2-way interaction strength on the off diagonal and individual variable importance on the diagonal.</caption>
  </center>

#### PDP
```{r, ke_r_pdp, out.width='100%', eval=FALSE}
pdpPairs(data = aqFinal, 
         fit =  model, 
         response = "Ozone", 
         nmax = 500, 
         gridSize = 10,         
         nIce = 100,
         predictFun = vi_fun)
```

```{r, echo = F,  out.width = '100%'}
knitr::include_graphics("https://raw.githubusercontent.com/AlanInglis/vivid/master/vignettes/vig/bm_r_pdp-1.png")
```
<center>
<caption><span id="fig2:fig2">Figure 2: </span>Generalized pairs partial dependence plot for a keras neural network regression fit.</caption>
</center>