Handling missing values (drop_na, fill) in R Programming

Introduction

Sometimes data has empty spots called missing values. We need to fix or remove them to keep our data clean and useful.

When you want to remove rows that have missing data before analysis.

When you want to fill missing spots with a specific value like zero or the last known value.

When preparing data for charts or models that can't handle missing values.

When cleaning survey or experiment data that has incomplete answers.

Syntax

R Programming

library(tidyr)
drop_na(data, columns)

library(tidyr)
fill(data, columns, .direction = c("down", "up"))

drop_na() removes rows with missing values in specified columns.

fill() fills missing values by carrying the previous or next non-missing value.

Examples

Remove rows where column1 has missing values.

R Programming

drop_na(df, column1)

Remove rows where any column has missing values.

R Programming

drop_na(df)

Fill missing values in column2 by copying the last known value downwards.

R Programming

fill(df, column2, .direction = "down")

Fill missing values in column3 by copying the next known value upwards.

R Programming

fill(df, column3, .direction = "up")

Sample Program

This program shows how to remove rows with missing score values and how to fill missing grade values by copying the last known grade downwards.

R Programming

library(tidyr)
library(dplyr)

# Create example data frame with missing values
df <- data.frame(
  id = 1:5,
  score = c(10, NA, 15, NA, 20),
  grade = c("A", NA, "B", NA, "C")
)

print("Original data frame:")
print(df)

# Remove rows with missing values in 'score'
df_no_na <- drop_na(df, score)
print("\nData frame after drop_na on 'score':")
print(df_no_na)

# Fill missing values in 'grade' by carrying last observation forward
df_filled <- fill(df, grade, .direction = "down")
print("\nData frame after fill on 'grade' (down):")
print(df_filled)

Important Notes

Remember to load tidyr and dplyr packages before using drop_na() and fill().

drop_na() removes whole rows, so use carefully if you want to keep some data.

fill() only works on columns with missing values and copies existing values up or down.

Summary

drop_na() removes rows with missing values in chosen columns.

fill() fills missing values by copying nearby known values.

Use these to clean data and prepare it for analysis or visualization.

Practice

1. What does the drop_na() function do in R's tidyr package?

easy

A. Fills missing values with the mean of the column.

B. Replaces missing values with zeros.

C. Removes rows that contain missing values in specified columns.

D. Sorts the data frame by missing values.

Solution

1. Step 1: Understand the purpose of drop_na()

   The function is designed to remove rows that have missing (NA) values in specified columns or in the entire data frame if no columns are specified.

2. Step 2: Compare with other options

   Options A and C describe filling missing values, which is done by fill() or other functions, not drop_na(). Sorts the data frame by missing values. sorts the data frame by missing values, which is unrelated.

3. Final Answer:

   Removes rows that contain missing values in specified columns. -> Option C

4. Quick Check:

   drop_na() removes rows with NA [OK]

Hint: Remember: drop_na drops rows with NA, fill replaces NA [OK]

Common Mistakes:

• Confusing drop_na() with fill()
• Thinking drop_na() fills missing values
• Assuming drop_na() sorts data

2. Which of the following is the correct syntax to fill missing values in a data frame df using the fill() function from tidyr for the column score?

easy

A. df.fill(score, direction = "down")

B. fill(df, score, direction = "down")

C. fill(score, df, .direction = "down")

D. df %>% fill(score, .direction = "down")

Solution

1. Step 1: Recall the pipe operator usage

   The fill() function is typically used with the pipe operator %>% to pass the data frame as the first argument.

2. Step 2: Check the argument order and syntax

   df %>% fill(score, .direction = "down") correctly uses df %>% fill(score, .direction = "down"). Options B and C have wrong syntax or argument order. df.fill(score, direction = "down") uses incorrect syntax not valid in R.

3. Final Answer:

   df %>% fill(score, .direction = "down") -> Option D

4. Quick Check:

   Use pipe %>% before fill() with column name [OK]

Hint: Use pipe %>% before fill() and specify column [OK]

Common Mistakes:

• Passing data frame as second argument
• Using dot direction without dot prefix
• Using Python-like method syntax

3. Given the data frame df below:

df <- data.frame(id = 1:5, score = c(10, NA, 15, NA, 20))
df_filled <- df %>% fill(score, .direction = "down")
print(df_filled)

What will be the output of print(df_filled)?

medium

A. id score
1 1 10
2 2 10
3 3 15
4 4 15
5 5 20

B. id score
1 1 10
2 2 NA
3 3 15
4 4 NA
5 5 20

C. id score
1 1 10
2 2 15
3 3 15
4 4 20
5 5 20

D. id score
1 1 NA
2 2 NA
3 3 15
4 4 15
5 5 20

Solution

1. Step 1: Understand fill() with .direction = "down"

   The function fills missing values by copying the last known non-NA value downward to replace NA values.

2. Step 2: Apply fill() to the data

   Row 2's NA is replaced by 10 (from row 1), row 4's NA is replaced by 15 (from row 3). Other values remain unchanged.

3. Final Answer:

   id score 1 1 10 2 2 10 3 3 15 4 4 15 5 5 20 -> Option A

4. Quick Check:

   fill() copies last known value down [OK]

Hint: fill() with down copies last known value to NA below [OK]

Common Mistakes:

• Assuming fill() replaces NA with next value
• Thinking fill() leaves NA unchanged
• Confusing fill() direction

4. Consider the following code snippet:

library(tidyr)
df <- data.frame(x = c(1, NA, 3), y = c(NA, 2, 3))
df_clean <- drop_na(df, x)
print(df_clean)

What is the output of print(df_clean)?

medium

A. x y
1 1 NA
3 3 3

B. x y
2 NA 2
3 3 3

C. x y
1 1 NA
2 NA 2
3 3 3

D. x y
3 3 3

Solution

1. Step 1: Understand drop_na(df, x)

   This removes rows where column x has NA values. Row 2 has NA in x, so it is removed.

2. Step 2: Identify remaining rows

   Rows 1 and 3 remain. Row 1 has x=1, y=NA. Row 3 has x=3, y=3.

3. Final Answer:

   x y 1 1 NA 3 3 3 -> Option A

4. Quick Check:

   drop_na(df, x) removes rows with NA in x [OK]

Hint: drop_na(df, col) removes rows with NA in that column [OK]

Common Mistakes:

• Removing rows with NA in all columns instead of specified
• Confusing drop_na() with fill()
• Expecting NA values to be replaced

5. You have a data frame df with columns time and temperature, where some temperature values are missing (NA). You want to fill missing temperatures by carrying the last known temperature forward, but only within each day (grouped by time). Which code correctly achieves this?

hard

A. df %>% fill(temperature, .direction = "up") %>% group_by(time)

B. df %>% group_by(time) %>% fill(temperature, .direction = "down")

C. df %>% fill(temperature, .direction = "down") %>% group_by(time)

D. df %>% group_by(time) %>% drop_na(temperature)

Solution

1. Step 1: Group data by time

   Grouping by time ensures fill() works within each day separately.

2. Step 2: Use fill() with .direction = "down"

   Filling downward copies last known temperature forward within each group.

3. Step 3: Check other options

   df %>% fill(temperature, .direction = "down") %>% group_by(time) fills before grouping, so fill applies to whole data ignoring groups. df %>% fill(temperature, .direction = "up") %>% group_by(time) fills upward, which is opposite of requirement. df %>% group_by(time) %>% drop_na(temperature) drops rows with NA instead of filling.

4. Final Answer:

   df %>% group_by(time) %>% fill(temperature, .direction = "down") -> Option B

5. Quick Check:

   Group then fill down to fill within groups [OK]

Hint: Group first, then fill down to fill missing within groups [OK]

Common Mistakes:

• Filling before grouping
• Using fill direction up instead of down
• Dropping rows instead of filling