---
title: "Stat. 652 - NHANES lm"
author: "Prof. Eric A. Suess"
format: 
  html:
    embed-resources: true
---

Here we try using Tidymodels and Recipe to fit a prediction, the lm model.

Note the need to impute values because of the high percentage of missing values.

For the glm model we use all of the features in the data set.

```{r}
library(pacman)

p_load(NHANES, tidyverse, tidymodels, skimr, naniar)
```

### 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}
skim(NHANES)
```



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

Summarize the y-variable.  This is the Null Model.

```{r}
NHANES_SleepHrsNight |> group_by(SleepHrsNight) |>
  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_SleepHrsNight_split <- initial_split(NHANES_SleepHrsNight, prop = 0.8)
NHANES_SleepHrsNight_split
```

Training data.

```{r}
NHANES_SleepHrsNight_split |>
  training() 
```

```{r}
NHANES_SleepHrsNight_split |>
  training() |>
  vis_miss()
```

```{r}
NHANES_SleepHrsNight_split |>
  testing() 
```

```{r}
NHANES_SleepHrsNight_split |>
  testing() |>
  vis_miss()
```

Create the *recipe()* for applying the preprocessing using *prep()*.  Note the use of *step_nzv()*, which removes any columns that have very low variability, and the use of the *step_impute_median()* function, which fills in the cells that are missing with the median of the column.

```{r}
NHANES_SleepHrsNight_recipe <- training(NHANES_SleepHrsNight_split) |>
  recipe(SleepHrsNight ~ .) |>
  step_nzv(all_predictors()) |>
  step_impute_knn(all_predictors()) |>
  step_corr(all_numeric()) |>
  prep()

summary(NHANES_SleepTrouble_recipe)

tidy(NHANES_SleepTrouble_recipe)
```

Note that the variable SleepHrsNight is moved to the last column of the dataset.

We "bake" the receipe into the Training data to see what the preprocesing does to the data.  We do the same for the Testing data.

```{r}
NHANES_SleepHrsNight_training <- NHANES_SleepHrsNight_recipe |>
  bake(training(NHANES_SleepHrsNight_split)) 

NHANES_SleepHrsNight_training
```


```{r}
NHANES_SleepHrsNight_testing <- NHANES_SleepHrsNight_recipe |>
  bake(testing(NHANES_SleepHrsNight_split)) 

NHANES_SleepHrsNight_testing
```

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

Setup the models.

```{r}

NHANES_SleepHrsNight_lm <- linear_reg() |> 
  set_engine("lm") |>
  set_mode("regression") |>
  fit(SleepHrsNight ~ ., data = NHANES_SleepHrsNight_training)

```


```{r}
predict(NHANES_SleepHrsNight_lm, NHANES_SleepHrsNight_testing)
```

```{r}
NHANES_SleepHrsNight_lm |>
  predict(NHANES_SleepHrsNight_testing) |>
  bind_cols(NHANES_SleepHrsNight_testing) 
```

### Step 4: Evaluate the models.

```{r}
NHANES_SleepHrsNight_lm |>
  predict(NHANES_SleepHrsNight_testing) |>
  bind_cols(NHANES_SleepHrsNight_testing) |>
  metrics(truth = SleepHrsNight, estimate = .pred)
```


```{r}
NHANES_SleepHrsNight_lm |>
  predict(NHANES_SleepHrsNight_testing) |>
  bind_cols(NHANES_SleepHrsNight_testing) |>
  rmse(truth = SleepHrsNight, estimate = .pred)

```


### Step 5: Improve the model.

GLM model using regularization.

Setup the models.

```{r}

NHANES_SleepHrsNight_glm <- linear_reg(penalty = 0.001, mixture = 0.5) |> 
  set_engine("glmnet") |>
  set_mode("regression") |>
  fit(SleepHrsNight ~ ., data = NHANES_SleepHrsNight_training)

```


```{r}
predict(NHANES_SleepHrsNight_glm, NHANES_SleepHrsNight_testing)
```

```{r}
NHANES_SleepHrsNight_glm |>
  predict(NHANES_SleepHrsNight_testing) |>
  bind_cols(NHANES_SleepHrsNight_testing) 
```


```{r}
NHANES_SleepHrsNight_glm |>
  predict(NHANES_SleepHrsNight_testing) |>
  bind_cols(NHANES_SleepHrsNight_testing) |>
  metrics(truth = SleepHrsNight, estimate = .pred)
```


```{r}
NHANES_SleepHrsNight_glm |>
  predict(NHANES_SleepHrsNight_testing) |>
  bind_cols(NHANES_SleepHrsNight_testing) |>
  rmse(truth = SleepHrsNight, estimate = .pred)

```

Setup the models.

```{r}

NHANES_SleepHrsNight_ranger <- rand_forest(trees = 1000) |> 
  set_engine("ranger") |>
  set_mode("regression") |>
  fit(SleepHrsNight ~ ., data = NHANES_SleepHrsNight_training)

```


```{r}
predict(NHANES_SleepHrsNight_ranger, NHANES_SleepHrsNight_testing)
```

```{r}
NHANES_SleepHrsNight_ranger |>
  predict(NHANES_SleepHrsNight_testing) |>
  bind_cols(NHANES_SleepHrsNight_testing) 
```


```{r}
NHANES_SleepHrsNight_ranger |>
  predict(NHANES_SleepHrsNight_testing) |>
  bind_cols(NHANES_SleepHrsNight_testing) |>
  metrics(truth = SleepHrsNight, estimate = .pred)
```


```{r}
NHANES_SleepHrsNight_ranger |>
  predict(NHANES_SleepHrsNight_testing) |>
  bind_cols(NHANES_SleepHrsNight_testing) |>
  rmse(truth = SleepHrsNight, estimate = .pred)

```