Accessing columns ($, []) in R Programming - Time & Space Complexity

Time Complexity: Accessing columns ($, [])

O(m)

Understanding Time Complexity

We want to understand how the time to get a column from a data frame changes as the data grows.

How does the cost of accessing columns with $ or [] grow when the data frame gets bigger?

Scenario Under Consideration

Analyze the time complexity of the following code snippet.

# Create a data frame with n rows and 3 columns
n <- 1000
df <- data.frame(a = 1:n, b = rnorm(n), c = letters[(1:n %% 26) + 1])

# Access column 'b' using $
col1 <- df$b

# Access column 'b' using []
col2 <- df["b"]
    

This code creates a data frame and accesses one column in two ways: with $ and with [].

Identify Repeating Operations

Identify the loops, recursion, array traversals that repeat.

• Primary operation: Extracting a single column from the data frame.
• How many times: The operation happens once per access, but internally it may scan column names.

How Execution Grows With Input

When the data frame has more rows, accessing a column still takes about the same time because it just points to that column's data.

Input Size (n)	Approx. Operations
10	Few operations to find and return the column
100	Similar few operations, no big change
1000	Still about the same operations, just returning a bigger vector

Pattern observation: The time to access a column grows very little with more rows; it mainly depends on the number of columns.

Final Time Complexity

Time Complexity: O(m)

This means the time to access a column grows mostly with the number of columns (m), not the number of rows.

Common Mistake

[X] Wrong: "Accessing a column takes longer if the data frame has more rows."

[OK] Correct: Accessing a column just returns a reference or copy of that column's data, so the number of rows does not affect the search time, only the size of the returned data.

Interview Connect

Knowing how data frame column access scales helps you write efficient data code and explain your choices clearly in interviews.

Self-Check

"What if we accessed multiple columns at once using df[c('a','b')]? How would the time complexity change?"