---
title: "R Notebook"
output: html_notebook
---



```{r}
library(pacman)

p_load(NHANES, tidyverse, tidymodels, discrim)
```

### Step 1: Access the data.

```{r}
data(NHANES)

NHANES

```

### Step 2: Clean the data.  Prepare the data and make a recipe for applying the steps needed to preprocess the data.

First drop all of the rows where the y-variable SleepTrouble is missing.

```{r}
NHANES_SleepTrouble <- NHANES %>% select(-ID, -SleepHrsNight) %>%
  select( SleepTrouble, everything()) %>%
  drop_na(SleepTrouble)
NHANES_SleepTrouble

```

Summarize the y-variable.

```{r}
NHANES_SleepTrouble %>% group_by(SleepTrouble) %>%
  summarize(n = n()) %>%
  mutate(freq = n / sum(n))
```

Make the first split with 80% of the data being in the trainning data set.

```{r}
NHANES_SleepTrouble_split <- initial_split(NHANES_SleepTrouble, prop = 0.8)
NHANES_SleepTrouble_split
```

Trainning data.

```{r}
NHANES_SleepTrouble_split %>%
  training() 
```

```{r}
NHANES_SleepTrouble_split %>%
  training() %>%
  vis_miss()
```


Create the recipe for applying the preprocessing.  Note the use of step_nzv(), which removes any columns that have very low variability, the use of the step_knnimpute() function, which fills in the cells that are missing with the median of the column, and tghe use of the step_corr() function, which removes highly correlated input features.


```{r}
NHANES_SleepTrouble_recipe <- training(NHANES_SleepTrouble_split) %>%
  recipe(SleepTrouble ~ .) %>%
  step_nzv(all_predictors()) %>%
  step_knnimpute(all_predictors()) %>%
  step_corr(all_numeric()) %>%
  prep()

summary(NHANES_SleepTrouble_recipe)

tidy(NHANES_SleepTrouble_recipe)
```


```{r}
NHANES_SleepTrouble_testing <- NHANES_SleepTrouble_recipe %>%
  bake(testing(NHANES_SleepTrouble_split)) 

NHANES_SleepTrouble_testing
```


```{r}
NHANES_SleepTrouble_training <- juice(NHANES_SleepTrouble_recipe)

NHANES_SleepTrouble_training
```

### Step 3: Training a model on the data

Setup the models.

```{r}

NHANES_SleepTrouble_glm <- logistic_reg() %>% 
  set_engine("glm") %>%
  set_mode("classification") %>%
  fit(SleepTrouble ~ ., data = NHANES_SleepTrouble_training)

```


```{r}
predict(NHANES_SleepTrouble_glm, NHANES_SleepTrouble_testing)
```

```{r}
NHANES_SleepTrouble_glm %>%
  predict(NHANES_SleepTrouble_testing) %>%
  bind_cols(NHANES_SleepTrouble_testing) 
```


### Step 4: Evaluate the models.


```{r}
NHANES_SleepTrouble_glm %>%
  predict(NHANES_SleepTrouble_testing) %>%
  bind_cols(NHANES_SleepTrouble_testing) %>%
  metrics(truth = SleepTrouble, estimate = .pred_class)
```


```{r}
NHANES_SleepTrouble_glm %>%
  predict(NHANES_SleepTrouble_testing) %>%
  bind_cols(NHANES_SleepTrouble_testing) %>%
  conf_mat(truth = SleepTrouble, estimate = .pred_class)

```


```{r}
NHANES_SleepTrouble_glm %>%
  predict(NHANES_SleepTrouble_testing, type = "prob") %>%
  bind_cols(NHANES_SleepTrouble_testing) %>%
  roc_curve(SleepTrouble, .pred_Yes) %>%
  ggplot(aes(x = 1 - specificity, y = sensitivity)) +
  geom_path() +
  geom_abline(lty = 3) +
  coord_equal() 
```


```{r}
NHANES_SleepTrouble_glm %>%
  predict(NHANES_SleepTrouble_testing, type = "prob") %>%
  bind_cols(NHANES_SleepTrouble_testing) %>%
  roc_auc(SleepTrouble, .pred_Yes) 
```


### Step 5: Improve the model.

GLM model using regularization.

```{r}

NHANES_SleepTrouble_glmnet <- logistic_reg(penalty = 0.001, mixture = 0.5) %>% 
  set_engine("glmnet") %>%
  set_mode("classification") %>%
  fit(SleepTrouble ~ ., data = NHANES_SleepTrouble_training)

summary(NHANES_SleepTrouble_glmnet)

```


```{r}
predict(NHANES_SleepTrouble_glmnet, NHANES_SleepTrouble_testing)
```

```{r}
NHANES_SleepTrouble_glmnet %>%
  predict(NHANES_SleepTrouble_testing) %>%
  bind_cols(NHANES_SleepTrouble_testing) 
```


```{r}
NHANES_SleepTrouble_glmnet %>%
  predict(NHANES_SleepTrouble_testing) %>%
  bind_cols(NHANES_SleepTrouble_testing) %>%
  metrics(truth = SleepTrouble, estimate = .pred_class)
```


```{r}
NHANES_SleepTrouble_glmnet %>%
  predict(NHANES_SleepTrouble_testing) %>%
  bind_cols(NHANES_SleepTrouble_testing) %>%
  conf_mat(truth = SleepTrouble, estimate = .pred_class)

```


```{r}
NHANES_SleepTrouble_glmnet %>%
  predict(NHANES_SleepTrouble_testing, type = "prob") %>%
  bind_cols(NHANES_SleepTrouble_testing) %>%
  roc_curve(SleepTrouble, .pred_Yes) %>%
  ggplot(aes(x = 1 - specificity, y = sensitivity)) +
  geom_path() +
  geom_abline(lty = 3) +
  coord_equal() 
```



```{r}
NHANES_SleepTrouble_glmnet %>%
  predict(NHANES_SleepTrouble_testing, type = "prob") %>%
  bind_cols(NHANES_SleepTrouble_testing) %>%
  roc_auc(SleepTrouble, .pred_Yes) 
```

Naive Bayes model using regularization.

```{r}

NHANES_SleepTrouble_nb <- naive_Bayes(smoothness = .1) %>% 
  set_engine("klaR") %>%
  set_mode("classification") %>%
  fit(SleepTrouble ~ ., data = NHANES_SleepTrouble_training)

summary(NHANES_SleepTrouble_nb)

```


```{r}
predict(NHANES_SleepTrouble_nb, NHANES_SleepTrouble_testing)
```

```{r}
NHANES_SleepTrouble_nb %>%
  predict(NHANES_SleepTrouble_testing) %>%
  bind_cols(NHANES_SleepTrouble_testing) 
```


```{r}
NHANES_SleepTrouble_nb %>%
  predict(NHANES_SleepTrouble_testing) %>%
  bind_cols(NHANES_SleepTrouble_testing) %>%
  metrics(truth = SleepTrouble, estimate = .pred_class)
```


```{r}
NHANES_SleepTrouble_nb %>%
  predict(NHANES_SleepTrouble_testing) %>%
  bind_cols(NHANES_SleepTrouble_testing) %>%
  conf_mat(truth = SleepTrouble, estimate = .pred_class)

```


```{r}
NHANES_SleepTrouble_nb %>%
  predict(NHANES_SleepTrouble_testing, type = "prob") %>%
  bind_cols(NHANES_SleepTrouble_testing) %>%
  roc_curve(SleepTrouble, .pred_Yes) %>%
  ggplot(aes(x = 1 - specificity, y = sensitivity)) +
  geom_path() +
  geom_abline(lty = 3) +
  coord_equal() 
```



```{r}
NHANES_SleepTrouble_nb %>%
  predict(NHANES_SleepTrouble_testing, type = "prob") %>%
  bind_cols(NHANES_SleepTrouble_testing) %>%
  roc_auc(SleepTrouble, .pred_Yes) 
```