presentation and solution 4A&4b

Author: helenewegener

presentations/presentation4A_conditionsNloops.qmd

@@ -52,7 +52,6 @@ Now we have three possible outcomes:
 
 
 ```{r}
-
 #try with different values for num2
 num2 <- 10
 
@@ -67,6 +66,51 @@ if (num1 > num2){
 print(statement)
 ```
 
+### And and or operations 
+
+Check values of num1 and num2
+```{r}
+num1
+num2
+
+# num1 <- 14
+# num2 <- 12
+```
+
+You can give multiple conditions and check if both of them are true with the `&` (and-operation).
+```{r}
+if (num1 < 10 & num2 < 10) {
+  print('Both numbers are lower than 10')
+} else {
+  print('Both numbers are not lower than 10')
+}
+```
+
+You can also check if either one or the other is true with the `|` (or-operation).  
+```{r}
+if (num1 < 10 | num2 < 10) {
+  print('One or both of the numbers are lower than 10.')
+} else {
+  print('None of the numbers are not lower than 10')
+}
+```
+
+If you add more conditions, paraphrases can be useful. 
+```{r}
+# num1 <- 8
+# num2 <- 5
+
+num3 <- 10
+```
+
+```{r}
+if ((num1 < 10 | num2 < 10) & num3 == 10) {
+  print('Yes')
+} else {
+  print('No')
+}
+```
+
 ## For-loops
 
 ### Defining a for loop
@@ -115,14 +159,10 @@ THIS_VARIABLE
 
 ### Loop control
 
-There are two loop control statements we can use to 
-
-* jump to the next iteration: `next`
-* end the loop before finishing: `break`
+There are two loop control statements we can use to: `next` and `break`
 
+`next` jumps to the next iteration. Here, we print every element in list1 and when the element is 'hello' we jump to the next iteration.
 ```{r}
-#example for next
-
 for (element in list1) {
   if(element == 'hello'){
     next
@@ -132,8 +172,8 @@ for (element in list1) {
 }
 ```
 
+ `break` ends the loop before finishing. Here, we print every element in list1 and when the element is 'hello' we break (end) the loop.
 ```{r}
-#example for break
 for (element in list1) {
   if(element == 'hello'){
     break
@@ -192,6 +232,68 @@ for (i in 1:nrow(my_df)) {
 
 ```
 
+### Plotting in loops
+Create data
+```{r}
+plot_data_1 <- tibble(Name = c('Marie', 'Marie', 'Emma', 'Sofie', 'Sarah', 'Sofie', 'Hannah', 'Lise', 'Emma'),
+                      Class = c('1.A', '1.A', '1.A', '1.A', '1.B', '1.B', '1.B', '1.C', '1.C'),
+                      Food = c('Lasagna', 'Pizza', 'Pizza', 'Bruger', 'Lasagna', 'Lasagna', 'Lasagna', 'Burger', 'Lasagna'),
+                      Age = c(6, 6, 6, 6, 6, 5, 7, 6, 6))
+
+head(plot_data_1, n = 2)
+```
+Barplot of each variable.
+```{r}
+ggplot(plot_data_1, 
+       aes(x = Name)) + 
+  geom_bar()
+
+ggplot(plot_data_1, 
+       aes(x = Class)) + 
+  geom_bar()
+
+# and so on ...
+```
+
+Let's do it in a for loop!
+
+First, let's check that the variables we are interested in are iterated correctly. 
+```{r}
+for (col in colnames(plot_data_1)){
+  print(col)
+}
+```
+
+Great! Now, let's add the plot function to our for loop.
+```{r}
+for (col in colnames(plot_data_1)){
+  p <- ggplot(plot_data_1, 
+       aes(x = col)) + 
+  geom_bar()
+  
+  print(p)
+}
+```
+
+That was not what we wanted...
+
+Wrap the `!!sym()` function around the string formatted column name when passing it in aesthetic.
+
+* `sym()` turns a string into a symbol (column reference)
+* `!!` unquotes the symbol to use it in `aes()`
+
+```{r}
+for (col in colnames(plot_data_1)){
+  p <- ggplot(plot_data_1, 
+       aes(x = !!sym(col))) + 
+  geom_bar()
+  
+  print(p)
+}
+```
+
+`!!sym(col)` should also be used for other tidyverse operations (`filter`, `select`, ...) where you pass the column names in string format.
+
 
 ### If-else in loops
 

presentations/presentation4B_functions.qmd

@@ -106,9 +106,9 @@ if (age >= 18){
 Or we can choose to execute our function once for every element of an iterable, e.g. every row in a dataframe:
 
 ```{r}
-df <- data.frame(row.names = 1:5, 
-                 age = c(45, 16, 31, 56, 19), 
-                 weight_kg = c(85, 65, 100, 45, 76), 
+df <- data.frame(row.names = 1:5,
+                 age = c(45, 16, 31, 56, 19),
+                 weight_kg = c(85, 65, 100, 45, 76),
                  height_m = c(1.75, 1.45, 1.95, 1.51, 1.89))
 
 df
@@ -118,13 +118,13 @@ Print ID, weight, and height of all individuals.
 
 ```{r}
 for (id in rownames(df)){
-  
+
   weight <- df[id, 'weight_kg']
-  
+
   height <- df[id, 'height_m']
-  
+
   print(c(id, weight, height))
-  
+
 }
 ```
 
@@ -200,6 +200,36 @@ Have a look at the data frame.
 df
 ```
 
+## Plotting in functions
+
+Define function that creates boxplot
+```{r}
+my_boxplot <- function(dataframe, variable = ''){
+  
+  p <- ggplot(data = dataframe, 
+         aes(y = !!sym(variable))) + # Use variable as column reference
+  geom_boxplot(color = 'blue') + 
+  theme_bw() + 
+  labs(title = paste('Boxplot of', variable)) # Use variable as string
+  
+  return(p)
+  
+}
+```
+
+Look at column names of df
+```{r}
+colnames(df)
+colnames(df)[1]
+```
+
+Run function on age. 
+```{r}
+my_boxplot(dataframe = df, colnames(df)[1])
+```
+
+
+
 ## Error handling in user-defined functions
 
 Currently our BMI function accepts all kinds of inputs. However, what happens if we give a negative weight?

program_draft.txt

@@ -0,0 +1,23 @@
+DAY 1
+(30 minutes)    Introduction
+(30 minutes)    Exercise 0: Getting started
+(45 minutes)    Presentation 1: Data Cleanup & Summary Statistics
+(45-60 minutes) Exercise 1: Data Cleanup and Summary Statistics
+(45 minutes)    Presentation 2: Data Transformation and Integration
+(60 minutes)    Exercise 2: Data Transformation and Integration
+
+DAY 2
+(XX minutes)    Presentation 3 - Exploratory Data Analysis (EDA)
+(XX minutes)    Exercise 3, 3A, 3B ???
+(45 minutes)    Presentation 4A: Scripting in R - Conditions and For-loops
+(60 minutes)    Exercise 4A: Scripting in R - Conditions and For-loops
+(30 minutes)    Presentation 4B: Scripting in R - Functions
+(60 minutes)    Exercise 4B: Scripting in R - Functions
+
+DAY 3 
+(XX minutes)    Presentation 5A: Models and Model Evaluation in R
+(XX minutes)    Exercise 5A: Models and Model Evaluation in R
+(XX minutes)    Presentation 5B: Models and Model Evaluation in R
+(XX minutes)    Exercise 5B: Models and Model Evaluation in R
+
+

solutions/solution4A.qmd

@@ -106,6 +106,7 @@ for (i in 1:10) {
   } else {
     print('Normal blood pressure')
   } 
+  
 }
 ```
 
@@ -144,4 +145,102 @@ for (i in 1:10) {
 
 ```
 
+7. Do the same as above but instead of printing the risk status, append it to a list. Start by initiating an empty list. 
+
+```{r}
+# Initiate list
+risk_status <- list()
+```
+
+```{r}
+for (i in 1:nrow(diabetes_glucose)) {
+  Smoker <- diabetes_glucose$Smoker[i]
+  BMI <- diabetes_glucose$BMI[i]
+  
+  #skip rows where either of the values is NA
+  if (is.na(Smoker) | is.na(BMI)){
+    next
+  }
+    
+  if (Smoker == 'Smoker' & BMI > 35){
+    risk_status <- append(risk_status, 'High risk')
+  } else if (Smoker == 'Smoker' | BMI > 35) {
+    risk_status <- append(risk_status, 'Moderate risk')
+  } else {
+    risk_status <- append(risk_status, 'Low risk')
+  }
+}
+
+```
+
+```{r}
+risk_status %>% head()
+```
+
+8. Check the length of the list. Is it as expected?
+
+Since we looped through all the rows in the `diabetes_glucose` dataframe, the list should be as long as there are row in the dataframe. 
+```{r}
+length(risk_status)
+```
+
+```{r}
+nrow(diabetes_glucose)
+```
+
+
+9. Add the list as a column in the `diabetes_glucose` data frame. 
+
+```{r}
+diabetes_glucose$risk_status <- risk_status
+```
+
+```{r}
+diabetes_glucose %>% select(BMI, Smoker, risk_status) %>% unnest(risk_status)
+```
+
+10. Make a list of all the column names in `diabetes_glucose` that contain categorical variables. Make a for loop that goes through the list and prints a barplot for each of the categorical variables. 
+
+```{r}
+categorical <- list('Sex', 'Smoker', 'Diabetes', 'Married', 'Work')
+```
+
+```{r}
+for (var in categorical){
+  
+  p <- ggplot(diabetes_glucose, 
+         aes(x = !!sym(var))) + 
+    geom_bar() +
+    labs(title = paste('Barplot of', var))
+  
+  print(p)
+  
+}
+```
+
+10. Make a list of all the column names in `diabetes_glucose` that contain numeric variables. Make a for loop that goes through the list and prints a boxplot for each of the categorical variables. 
+
+```{r}
+head(diabetes_glucose)
+```
+
+```{r}
+numeric <- list('Age', 'BloodPressure', 'BMI', 'PhysicalActivity', 'Serum_ca2')
+```
+
+```{r}
+for (var in numeric){
+  
+  p <- ggplot(diabetes_glucose, 
+         aes(y = !!sym(var))) + 
+    geom_boxplot() + 
+    labs(title = paste('Boxplot of', var))
+  
+  print(p)
+  
+}
+```
+
+
+
 

solutions/solution4B.R

@@ -6,7 +6,7 @@ make_boxplot <- function(df, plot_column){
     stop('The column to plot must be numcerial.')
   }
 
-  p <- ggplot(df, aes(y = .data[[plot_column]])) +
+  p <- ggplot(df, aes(y = !!sym(plot_column))) +
     geom_boxplot(fill = "#03579A") +
     labs(title = paste("Boxplot of", plot_column)) + 
     theme_bw()