Recap :: Data Reshaping : Base R Tools

tidy data

Tidy data ¹ is a standardized way of organizing data values within a dataset. In tidy data:

• Each variable forms a column.
• Each observation forms a row.
• Each type of observational unit forms a table.

This structure makes it easier to manipulate, analyze, and visualize data.

Pipes |> or %>%

• |> or %>% operators allow you to pass the output of one function as the input to another function, making your code more readable and concise. Simply saying, chain multiple operations together.

Pipes |> or %>% …

Find the average weight of chicks by Diet on day 21, sorted in descending order of average weight using ChickWeight dataset.

# Step 1: Filter data
day21 <- ChickWeight[ChickWeight$Time == 21, ]

# Step 2: Compute mean weight per Diet
avg_weights <- aggregate(weight ~ Diet, data = day21, FUN = mean)

# Step 3: Sort by descending weight
avg_weights <- avg_weights[order(-avg_weights$weight), ]

avg_weights

  Diet   weight
3    3 270.3000
4    4 238.5556
2    2 214.7000
1    1 177.7500

library(dplyr)

ChickWeight %>%
  filter(Time == 21) %>%                      # 1. Keep only records at day 21
  group_by(Diet) %>%                          # 2. Group by Diet
  summarise(avg_weight = mean(weight)) %>%    # 3. Calculate average weight
  arrange(desc(avg_weight))                   # 4. Sort descending

# A tibble: 4 × 2
  Diet  avg_weight
  <fct>      <dbl>
1 3           270.
2 4           239.
3 2           215.
4 1           178.

Data Reshaping with dplyr

library(dplyr)
iris |> 
   filter(Sepal.Length > 5) |>                          # filter rows
   select(Sepal.Length, Petal.Length, Species) |>       # select columns
   mutate(SePa.Length=Sepal.Length+Petal.Length) |>     # add new column
   arrange(desc(SePa.Length)) |>                        # sort rows
   group_by(Species) |>                                 # group by Species
   summarise(mean = mean(SePa.Length), n = n())         # summarise

# A tibble: 3 × 3
  Species     mean     n
  <fct>      <dbl> <int>
1 setosa      6.82    22
2 versicolor 10.3     47
3 virginica  12.2     49

Using select() to pick columns

Using slice() to pick rows

Functional row selection with filter()

mutate() & transmute() to add or transform columns

arrange (sorting)

The arrange() function is used to sort data.

group_by + summarise (aggregation)

The group_by() function is used to group data by categorical variable(s). The summarise() function is used to summarize data.

distinct (deduplication)