Main Page

Server Info & Course Info

Contact


Loading......

Password

Section 1. Function Fundamentals

What is a Function?

Like a juicer machine:

  • Input: Whole fruits (arguments)
  • Processing: Juicing mechanism (function body)
  • Output: Fresh juice (return value)

Ready-made Functions: sum()

# Calculate total score
math_scores <- c(75, 88, 92, 65)
sum(math_scores)

# [1] 320

Ready-made Functions: mean()

# Find average score
mean(math_scores)
# [1] 80

# add element into a vector 
c(math_scores, NA)

What if we add NA value?
mean(c(math_scores, NA)) ➔ NA
mean(c(math_scores, NA), na.rm = TRUE) ➔ ??

Ready-made Functions: paste()

# Combine text
students <- c("Alice", "Bob", "Charlie")
paste(students, "got", math_scores)

# [1] "Alice got 75"   "Bob got 88"     "Charlie got 92"

Create Your First Function

c_to_f <- function(celsius) {
    fahrenheit <- celsius * 9/5 + 32
    return(fahrenheit)
   }

# function name: c_to_f
# claim a function: function
# parameter: celsius
# function body: curly braces {}
# return value: return()

# Test it!
c_to_f(0)   # 32°F (freezing point)
c_to_f(100) # 212°F (boiling point)

Question: f_to_c ?

convert a temperature from fahrenheit to celsius

Question: can you add some “print()” to describe the function of this function and the output of this function

Qustion: what kind of error will we meet?

Function Parameters: BMI Calculator

Calculate Body Mass Index

calculate_bmi <- function(weight_kg, height_m=1.7) {
    bmi <- weight_kg / (height_m ^ 2)
    return(bmi)
}

# Different ways to call
calculate_bmi(70)          # Uses default height
calculate_bmi(70, 1.8)     # Provides height
calculate_bmi(height_m=1.75, weight_kg=65) # Named parameters

Question: BMI and Height => Weight ?

“x^1/2” vs “x^(1/2)”

When Things Go Wrong

c_to_f("thirty")  # Oops!

Error in celsius * 9/5 : non-numeric argument to binary operator

Let’s Fix It Together!

# Improved version
c_to_f_safe <- function(celsius) {
  if(is.numeric(celsius)) {
    fahrenheit = celsius * 9/5 + 32
    return(fahrenheit)
  } else {
    "Please enter numbers only!"
  }
}

c_to_f_safe("30C") # Now shows friendly message

Section 2. Modular Programming in R

Why Modules Matter?

  • Like LEGO bricks:
    Reusable & Flexible

    # Without modules:  
calculate_some_score <- function(x) { ... }  
# Reused in 10 scripts = 10 copies to maintain

Creating “my_utils.R”

Create a text file with:

# File size converter
bytes_to_human <- function(bytes) {
  if (bytes >= 1024^3){ return(paste(round(bytes/1024^3,1), "GB")) }
  if (bytes >= 1024^2){ return(paste(round(bytes/1024^2,1), "MB")) }
  return(paste(bytes, "Bytes"))
}

# Fun password generator
generate_password <- function(length=4) {
    this_password= paste(sample(c(LETTERS, letters), length), collapse="")
    return(this_password)
}

Testing Our Utilities

bytes_to_human(2147483647)  # Returns "2.0 GB"
bytes_to_human(1048576)     # Returns "1.0 MB"

generate_password()         # "XkLb"
generate_password(6)        # "QwZyTa"

File management in Linux

Question: what’s the meaning of the following commands?

ls
pwd
mkdir my_r_modules
cd my_r_modules
nano my_utils.R

Loading Modules

Using source():

# Always check your working directory first!
getwd()

# Load utilities
source("my_utils.R")  # File must exist in working directory

# Now use functions:
bytes_to_human(5000000)

Source() Pitfalls

Common errors:

source("wrong_folder/my_utils.R")  # File not found
source("MY_UTILS.R")               # Case sensitivity in Linux!

Fix with:

source("/home/user/project/utils/my_utils.R")  # Full path

Module Benefits Recap

  1. Reuse code → Write once, use everywhere
  2. Organized projects → Easy to find components
  3. Team-friendly → Share modules with classmates

Section 3. Core Data Structures

1. Vectors vs Lists

What’s the difference?

# Vector (all elements same type)
math_scores <- c(85, 92, 78)
class(math_scores)  # numeric

# List (mixed types allowed)
student_record <- list(
  name = "Alice", 
  scores = c(85, 92, 78),
  passed = TRUE
)
class(student_record)  # list

Type Conversion Pitfalls

Dangerous automatic conversions:

# Problem scenario
blood_types <- c("A", "B", "O", 12)  # Number added
str(blood_types)  # All converted to character!

# Safer approach
blood_types <- c("A", "B", "O", as.character(12))

Type Issues

Correcting data types:

# Create problematic vector
mixed_data <- c(1, 2, "X", 4)

# Check type
class(mixed_data)  # character

# character cannot be changed to numeric value
corrected <- as.numeric(mixed_data)  # Warning about NAs
# Returns 1 2 NA 4

List Structure Demo

Let’s examine our student record:

# str() is used to get the structure of a variable
str(student_record)
# Output shows:
# List of 3
# $ name  : chr "Alice"
# $ scores: num [1:3] 85 92 78
# $ passed: logi TRUE

Hospital Patient Profile

Real-world list example:

patient <- list(
  id = "P2024001",
  tests = c(36.5, 120, 80),  # Temperature / Blood (systolic/diastolic) Pressure
  diagnosis = "hypertension"
)
# Access blood pressure
patient$tests[2:3]  # Returns 120 80

Updating List Elements

How to modify medical records:

# Add new test result
patient$tests <- c(patient$tests, 98.6)
# Change diagnosis
patient$diagnosis <- "stage 2 hypertension"

Question: build vector & list containg the name and hair color of three students.

Hash Tables in R

Creating medicine price dictionary:

# First install package
install.packages("hash")
library(hash)

drug_prices <- hash() 
drug_prices[["Aspirin"]] <- 5.99
drug_prices[["Lisinopril"]] <- 12.50

Hash Table Operations

Working with our medicine dictionary:

# Check existence
has.key("Aspirin", drug_prices)  # TRUE

# Get all drugs
keys(drug_prices)  # Shows "Aspirin" "Lisinopril"

# Get price
drug_prices[["Aspirin"]]  # Returns 5.99

Data Frames Basics

Creating class roster:

class_roster <- data.frame(
  student_id = c(101, 102, 103),
  name = c("Alice", "Bob", "Charlie"),
  age = c(20, 21, 19)
)
# Automatic conversion to factors?
str(class_roster)  # Check string handling

Data Frame

Understanding the structure:

# Column names
names(class_roster)  # student_id, name, age

# First 2 rows
head(class_roster, 2)
#   student_id  name age
# 1        101 Alice  20
# 2        102   Bob  21

Data Type Check

Essential sanity check:

# Check vector type
is.numeric(math_scores)  # TRUE

# Check list element type
is.character(patient$id)  # TRUE

# Check dataframe column
is.factor(class_roster$name)  # Depends on stringsAsFactors

Summary Table

Data structure cheat sheet: | Structure | Element Type | Key Feature | |————|————–|————————–| | Vector | Homogeneous | Fast operations | | List | Heterogeneous| Nested structures | | Data Frame | Tabular | Columns can have diff types | | Hash | Key-Value | Fast lookups |

Section 4. File Operations in R

What We’ll Learn Today

  1. Safe file operations
  2. Read/write CSV & Excel files
  3. Store data with RDS format
  4. Restore saved data

1.1 Setting Up Your Playground

Always start by setting working directory:

# Set your workspace
setwd("~/my_project")
getwd()  # Check current directory

# Better alternative (install.packages("here"))
library(here)
here()  # Shows safe project path

1.2 Path Safety Check

Avoid errors by checking paths first:

data_path <- here("data/weather.csv")

# Check if file exists
if(file.exists(data_path)) {
  print("All systems go!")
} else {
  dir.create("data")  # Create folder if missing
  print("Created data folder!")
}

2.1 Let’s Read Weather Data

Simple CSV reading example:

# Sample weather data
weather <- data.frame(
  Day = c("Mon", "Tue", "Wed"),
  Temp = c(22, 25, 18)
)
write.csv(weather, "weather.csv", row.names = FALSE)

write.csv(weather, "weather1.csv", row.names = FALSE, quote=FALSE)

write.table(weather, "weather2.csv", row.names = FALSE, quote=FALSE, sep=',')

write.table(weather, "weather3.csv", row.names = FALSE, col.names=FALSE, quote=FALSE, sep=',')


# Read it back
my_data <- read.csv("weather.csv")
head(my_data)

Question: write a table seperated by “\t” with rownames and colnames

2.2 CSV Reading Pro Tips

Handle special cases properly:

# Read CSV with custom settings
read.csv("weather.csv",
         stringsAsFactors = FALSE,  # Keep text as text!
         na.strings = c("NA", ""))  # Catch missing values

“Factor” in R is not covered in our class

use as.character(factor_variable) to handle it.

if you are interested in “factor”, please read more information online.

3.1 Excel Files Made Easy

Working with Excel files:

# First install package: install.packages("readxl")
library(readxl)

# Read students' health data
students <- read_excel("students.xlsx", sheet = "体检数据")
head(students)

4.1 Why Use RDS?

Save complex objects perfectly:

# Create big matrix
huge_matrix <- matrix(rnorm(1000000), nrow=1000)

# Save space and time!
saveRDS(huge_matrix, "big_data.rds")

4.2 RDS vs CSV

Test the difference yourself:

system.time(saveRDS(huge_matrix, "test.rds"))  # Binary
system.time(write.csv(huge_matrix, "test.csv"))  # Text

5.1 Restore Your Data

Loading saved RDS files:

# Load the matrix back
recovered_data <- readRDS("big_data.rds")

# Check if identical
identical(huge_matrix, recovered_data)  # Should be TRUE!

Checklist

✓ Always check file paths ✓ Use RDS for big datasets ✓ readxl for Excel files ✓ Test data after loading

Section 5. Hospital Data Management System

Today’s Case Study

Scenario: Simulate daily temperature monitoring in a hospital

Example dataset structure:

# patient_id | date       | temperature
#---------------------------------------
# P001      | 2023-01-01 | 36.5
# P002      | 2023-01-01 | 41.2
# P003      | 2023-01-01 | 37.8

Step 1: Reading CSV Data

Basic CSV file reading in R:

# Create sample data file
write.csv(data.frame(
  patient_id = c("P001", "P002", "P003"),
  date = rep("2023-01-01", 3),
  temperature = c(36.5, 41.2, 37.8)
), "temp_data.csv")

# Read data
hospital_data <- read.csv("temp_data.csv")
head(hospital_data)

Step 2: Handling Abnormal Values

Create a value-cleaning function:

clean_temps <- function(temp) {
  ifelse(temp < 35 | temp > 41, 
         NA,  # Mark extreme values as missing
         round(temp, 1))
}

# Apply to our data
hospital_data$cleaned_temp <- clean_temps(hospital_data$temperature)
print(hospital_data)

Step 3: Daily Reports with Lists

Store daily reports using lists:

daily_report <- list(
  date = "2023-01-01",
  patients = nrow(hospital_data),
  avg_temp = mean(hospital_data$cleaned_temp, na.rm = TRUE),
  alerts = sum(as.numeric(is.na(hospital_data$cleaned_temp)))
  )
  
print(str(daily_report))

Step 4: Quick Lookup with Hash Tables

Create a simple patient index:

# Create named vector (simple hash table)
patient_index <- setNames(
  as.list(hospital_data$cleaned_temp),
  hospital_data$patient_id
)

# Query patient P002
print(paste0("P001's temperature:", patient_index$P001))
print(paste0("P002's temperature:", patient_index$P002))

Step 5: Saving Reports with RDS

Save/load objects while preserving data types:

saveRDS(daily_report, "daily_report_20230101.rds")

# Later loading
loaded_report <- readRDS("daily_report_20230101.rds")
print(loaded_report$alerts)

Common Error: Missing File

What happens when file doesn’t exist:

missing_data <- read.csv("non_existent_file.csv")

Error Handling Demo

Add simple file existence check:

if(file.exists("data.csv")) {
  safe_data <- read.csv("data.csv")
} else {
  warning("File not found! Using empty dataset")
  safe_data <- data.frame()
}