minor slide updates and created student script

docs/03-supervised-modeling-process.Rmd

@@ -93,7 +93,7 @@ knitr::include_graphics("images/modeling_process.png")
 ]
 
 ---
-# Prerequisites
+# Prerequisites .red[`r anicon::faa("hand-point-right", color = "red", animate = "horizontal")` code chunk 1]
 
 .pull-left[
 
@@ -169,7 +169,7 @@ knitr::include_graphics("images/nope.png")
    - as $n \geq p$ (where *p* represents the number of features), larger samples sizes are often required to identify consistent signals in the features
 
 ---
-# Mechanics of data splitting
+# Mechanics of data splitting .red[`r anicon::faa("hand-point-right", color = "red", animate = "horizontal")` code chunk 2]
 
 Two most common ways of splitting data include:
 
@@ -219,7 +219,7 @@ ggplot(train, aes(x = price)) +
 
 ---
 class: yourturn
-# Your Turn!
+# Your Turn! .red[`r anicon::faa("hand-point-right", color = "red", animate = "horizontal")` code chunk 3]
 
 1. Use __rsample__ to split the Ames housing data (`ames`) and the Employee attrition data (`churn`) using stratified sampling and with a 80% split.
 
@@ -970,7 +970,7 @@ Let's put these pieces together and analyze the Ames housing data:
 
 
 ---
-# Putting the process together
+# Putting the process together .red[`r anicon::faa("hand-point-right", color = "red", animate = "horizontal")` code chunk 4]
 
 .scrollable90[
 .pull-left[

docs/03-supervised-modeling-process.html

@@ -73,7 +73,7 @@
 ]
 
 ---
-# Prerequisites
+# Prerequisites .red[<span>&lt;i class="fas  fa-hand-point-right faa-horizontal animated " style=" color:red;"&gt;&lt;/i&gt;</span> code chunk 1]
 
 .pull-left[
 
@@ -149,7 +149,7 @@
    - as `\(n \geq p\)` (where *p* represents the number of features), larger samples sizes are often required to identify consistent signals in the features
 
 ---
-# Mechanics of data splitting
+# Mechanics of data splitting .red[<span>&lt;i class="fas  fa-hand-point-right faa-horizontal animated " style=" color:red;"&gt;&lt;/i&gt;</span> code chunk 2]
 
 Two most common ways of splitting data include:
 
@@ -187,7 +187,7 @@
 
 ---
 class: yourturn
-# Your Turn!
+# Your Turn! .red[<span>&lt;i class="fas  fa-hand-point-right faa-horizontal animated " style=" color:red;"&gt;&lt;/i&gt;</span> code chunk 3]
 
 1. Use __rsample__ to split the Ames housing data (`ames`) and the Employee attrition data (`churn`) using stratified sampling and with a 80% split.
 
@@ -807,7 +807,7 @@
 
 
 ---
-# Putting the process together
+# Putting the process together .red[<span>&lt;i class="fas  fa-hand-point-right faa-horizontal animated " style=" color:red;"&gt;&lt;/i&gt;</span> code chunk 4]
 
 .scrollable90[
 .pull-left[

student-scripts/03-supervised-modeling-process.Rmd

@@ -0,0 +1,119 @@
+---
+title: "Supervised Modeling Process"
+output: html_notebook
+---
+
+# Prerequisites
+
+```{r slide-4}
+# Packages required
+library(rsample)
+library(caret)
+library(tidyverse)
+
+# Data required
+## ames data
+ames <- AmesHousing::make_ames()
+
+## attrition data
+churn <- rsample::attrition
+```
+
+
+# Mechanics of data splitting 
+
+Two most common ways of splitting data include:
+
+* simple random sampling: randomly select observations
+* stratified sampling: preserving distributions
+   - classification: sampling within the classes to preserve the 
+     distribution of the outcome in the training and test sets
+   - regression: determine the quartiles of the data set and sample within those
+      artificial groups
+
+```{r slide-8}
+set.seed(123) # for reproducibility
+split <- initial_split(diamonds, strata = "price", prop = 0.7)
+train <- training(split)
+test  <- testing(split)
+
+# Do the distributions line up? 
+ggplot(train, aes(x = price)) + 
+  geom_line(stat = "density", 
+            trim = TRUE) + 
+  geom_line(data = test, 
+            stat = "density", 
+            trim = TRUE, col = "red")
+```
+
+
+# Your Turn! 
+
+1. Use __rsample__ to split the Ames housing data (`ames`) and the Employee attrition data (`churn`) using stratified sampling and with a 80% split.
+2. Verify that the distribution between training and test sets are similar.
+
+```{r slide-9}
+# ames data
+set.seed(123)
+ames_split <- initial_split(ames, prop = _____, strata = "Sale_Price")
+ames_train <- training(_____)
+ames_test  <- testing(_____)
+
+# attrition data
+set.seed(123)
+churn_split <- initial_split(churn, prop = _____, strata = "Attrition")
+churn_train <- training(_____)
+churn_test  <- testing(_____)
+```
+
+
+# Putting the process together 
+
+Let's put these pieces together and analyze the Ames housing data:
+
+1. Split into training vs testing data
+2. Specify a resampling procedure
+3. Create our hyperparameter grid
+4. Execute grid search
+5. Evaluate performance
+
+___This grid search takes ~2 min___
+
+```{r slide-35}
+# 1. stratified sampling with the rsample package
+set.seed(123)
+split  <- initial_split(ames, prop = 0.7, strata = "Sale_Price")
+ames_train  <- training(split)
+ames_test   <- testing(split)
+
+# 2. create a resampling method
+cv <- trainControl(
+  method = "repeatedcv", 
+  number = 10, 
+  repeats = 5
+  )
+
+# 3. create a hyperparameter grid search
+hyper_grid <- expand.grid(k = seq(2, 26, by = 2))
+
+# 4. execute grid search with knn model
+#    use RMSE as preferred metric
+knn_fit <- train(
+  Sale_Price ~ ., 
+  data = ames_train, 
+  method = "knn", 
+  trControl = cv, 
+  tuneGrid = hyper_grid,
+  metric = "RMSE"
+  )
+
+# 5. evaluate results
+# print model results
+knn_fit
+
+# plot cross validation results
+ggplot(knn_fit$results, aes(k, RMSE)) + 
+  geom_line() +
+  geom_point() +
+  scale_y_continuous(labels = scales::dollar)
+```