Lernen Tidying Data with tidyr | Data Cleaning and Wrangling Essentials

Swipe um das Menü anzuzeigen

Definition

Tidy data is a way of organizing datasets so that each variable forms a column, each observation forms a row, and each type of observational unit forms a table. The principles behind tidy data ensure that your data is structured in a consistent and predictable way, making it easier to manipulate, analyze, and visualize.

When working with real-world data, you often encounter datasets that are not structured for easy analysis. The tidyr package in R provides powerful tools for reshaping and tidying your data. Two of the most important functions are pivot_longer() and pivot_wider(). These functions help you convert between wide and long data formats, which is essential for many analysis and visualization tasks.


              12345678910111213141516171819202122
            
# Simulate a wide dataset
library(tidyr)
library(dplyr)

wide_data <- tibble(
  id = 1:3,
  math_2022 = c(85, 90, 78),
  math_2023 = c(88, 92, 80),
  science_2022 = c(80, 85, 75),
  science_2023 = c(83, 89, 78)
)

# Convert wide data to long format using pivot_longer()
long_data <- wide_data %>%
  pivot_longer(
    cols = -id,
    names_to = c("subject", "year"),
    names_sep = "_",
    values_to = "score"
  )

print(long_data)

You use pivot_wider() when you want to spread long-format data into a wide format, often to make comparisons across columns or to prepare data for specific types of analysis. This is useful when each row in your dataset represents a single observation, but you want to see each variable or measurement as its own column.

Key situations to use pivot_wider():

Transforming long data so each measurement type becomes a separate column;
Making it easier to compare values across categories or time periods;
Preparing data for functions or visualizations that require wide format.

Example: If you have a dataset where each row records a student's score in a subject and year, you can use pivot_wider() to create separate columns for each subject and year combination, making side-by-side comparison straightforward.


              1234567891011121314151617181920
            
# Simulate a dataset with combined columns
library(tidyr)
library(dplyr)

data <- tibble(
  id = 1:3,
  info = c("A_2022", "B_2023", "C_2022"),
  score = c(90, 85, 88)
)

# Separate the 'info' column into 'group' and 'year'
separated_data <- data %>%
  separate(info, into = c("group", "year"), sep = "_")

# Unite 'group' and 'year' back into a single column
united_data <- separated_data %>%
  unite("info", group, year, sep = "_")

print(separated_data)
print(united_data)

Tidying data is a crucial step for preparing your datasets for analysis or visualization. Clean and well-structured data makes it easier to use statistical functions, create plots, and share results with others. Functions like pivot_longer(), pivot_wider(), separate(), and unite() from the tidyr package give you the flexibility to reshape your data as needed for your analysis goals.

Key use cases for tidying data: