R r‑statistics.co

EDA Exercises in R: 50 Real Practice Problems

Fifty exploratory data analysis exercises spanning inspection, distributions, missing values, outliers, relationships, and full EDA workflows. Hidden solutions, runnable code.

By Selva Prabhakaran  ·  Published May 11, 2026  ·  Last updated May 11, 2026
RRun this once before any exercise
library(dplyr) library(tidyr) library(ggplot2) library(tibble) library(e1071) library(ggridges) library(naniar)

  

  





Section 1. Data inspection (8 problems)



Exercise 1.1: Dimensions



Difficulty: Beginner. Get the row and column count of airquality.



Show solution



  

    
RInteractive R

    
dim(airquality)

    

      
      
    

    

  

  








Exercise 1.2: Glimpse columns



Difficulty: Beginner. Inspect column types and sample values of mtcars with glimpse.



Show solution



  

    
RInteractive R

    
glimpse(mtcars)

    

      
      
    

    

  

  








Exercise 1.3: Summary statistics



Difficulty: Beginner. Print summary of airquality and identify the column with the most NAs.



Show solution



  

    
RInteractive R

    
summary(airquality)
# Ozone has the most NAs (37)

    

      
      
    

    

  

  








Exercise 1.4: First 10 rows



Difficulty: Beginner. Inspect the first 10 rows of diamonds.



Show solution



  

    
RInteractive R

    
head(diamonds, 10)

    

      
      
    

    

  

  








Exercise 1.5: Class of each column



Difficulty: Intermediate. Get the class of each column of iris as a named vector.



Show solution



  

    
RInteractive R

    
sapply(iris, class)

    

      
      
    

    

  

  








Exercise 1.6: Number of distinct values per column



Difficulty: Intermediate. For diamonds, count distinct values per column.



Show solution



  

    
RInteractive R

    
diamonds |>
  summarise(across(everything(), n_distinct)) |>
  pivot_longer(everything())

    

      
      
    

    

  

  








Exercise 1.7: Range of each numeric column



Difficulty: Intermediate. Get min and max for each numeric column of iris.



Show solution



  

    
RInteractive R

    
iris |>
  summarise(across(where(is.numeric),
                   list(min = min, max = max),
                   .names = "{.col}_{.fn}"))

    

      
      
    

    

  

  








Exercise 1.8: Build a one-shot data profile



Difficulty: Advanced. For airquality, return a tibble with column name, class, NA count, distinct count.



Show solution



  

    
RInteractive R

    
tibble(
  column     = names(airquality),
  class      = sapply(airquality, function(x) class(x)[1]),
  n_na       = sapply(airquality, function(x) sum(is.na(x))),
  n_distinct = sapply(airquality, n_distinct)
)

    

      
      
    

    

  

  








Section 2. Distributions (10 problems)



Exercise 2.1: Histogram of mpg



Difficulty: Beginner. Histogram of mtcars$mpg with 15 bins.



Show solution



  

    
RInteractive R

    
ggplot(mtcars, aes(mpg)) + geom_histogram(bins = 15)

    

      
      
    

    

  

  








Exercise 2.2: Density curve



Difficulty: Beginner. Density curve of diamonds$price.



Show solution



  

    
RInteractive R

    
ggplot(diamonds, aes(price)) + geom_density()

    

      
      
    

    

  

  








Exercise 2.3: Boxplot per group



Difficulty: Intermediate. Boxplot of Sepal.Length by Species.



Show solution



  

    
RInteractive R

    
ggplot(iris, aes(Species, Sepal.Length)) + geom_boxplot()

    

      
      
    

    

  

  








Exercise 2.4: Overlapping densities



Difficulty: Intermediate. Overlapping density plot of Sepal.Length, colored by Species, with alpha.



Show solution



  

    
RInteractive R

    
ggplot(iris, aes(Sepal.Length, fill = Species)) +
  geom_density(alpha = 0.5)

    

      
      
    

    

  

  








Exercise 2.5: Quintiles



Difficulty: Intermediate. Compute the 20th, 40th, 60th, 80th percentiles of mtcars$mpg.



Show solution



  

    
RInteractive R

    
quantile(mtcars$mpg, c(0.2, 0.4, 0.6, 0.8))

    

      
      
    

    

  

  








Exercise 2.6: Skewness and kurtosis



Difficulty: Advanced. Compute skewness and kurtosis of diamonds$price.



Show solution



  

    
RInteractive R

    
e1071::skewness(diamonds$price)
e1071::kurtosis(diamonds$price)
# Right-skewed (positive); heavy-tailed (kurt > 3)

    

      
      
    

    

  

  








Exercise 2.7: Log-transform a skewed variable



Difficulty: Intermediate. Plot log(price) histogram and observe the difference.



Show solution



  

    
RInteractive R

    
ggplot(diamonds, aes(log(price))) + geom_histogram(bins = 30)

    

      
      
    

    

  

  








Exercise 2.8: Histograms by facet



Difficulty: Intermediate. Histogram of price faceted by cut.



Show solution



  

    
RInteractive R

    
ggplot(diamonds, aes(price)) +
  geom_histogram(bins = 30) +
  facet_wrap(~ cut)

    

      
      
    

    

  

  








Exercise 2.9: Empirical CDF



Difficulty: Advanced. Plot the empirical CDF of mtcars$mpg.



Show solution



  

    
RInteractive R

    
ggplot(mtcars, aes(mpg)) + stat_ecdf()

    

      
      
    

    

  

  








Exercise 2.10: Compare distribution shape across groups



Difficulty: Advanced. Use ridgeline plots (ggridges) for diamond price by cut.



Show solution



  

    
RInteractive R

    
ggplot(diamonds, aes(x = price, y = cut, fill = cut)) +
  ggridges::geom_density_ridges(alpha = 0.6)

    

      
      
    

    

  

  








Section 3. Missing data (6 problems)



Exercise 3.1: Count NAs



Difficulty: Beginner. Total NA count in airquality.



Show solution



  

    
RInteractive R

    
sum(is.na(airquality))

    

      
      
    

    

  

  








Exercise 3.2: NA per column



Difficulty: Intermediate. NAs per column, sorted desc.



Show solution



  

    
RInteractive R

    
airquality |>
  summarise(across(everything(), ~ sum(is.na(.x)))) |>
  pivot_longer(everything()) |>
  arrange(desc(value))

    

      
      
    

    

  

  








Exercise 3.3: NA per row



Difficulty: Intermediate. Add a n_na column per row to airquality.



Show solution



  

    
RInteractive R

    
airquality |>
  mutate(n_na = rowSums(is.na(across(everything()))))

    

      
      
    

    

  

  








Exercise 3.4: Drop incomplete rows



Difficulty: Beginner. Remove rows with any NA.



Show solution



  

    
RInteractive R

    
drop_na(airquality)

    

      
      
    

    

  

  








Exercise 3.5: Visualize NA pattern



Difficulty: Advanced. Use naniar::vis_miss to visualize the missingness pattern.



Show solution



  

    
RInteractive R

    
naniar::vis_miss(airquality)

    

      
      
    

    

  

  








Exercise 3.6: Mean impute and document



Difficulty: Intermediate. Impute Ozone NAs with the column mean and add a flag column.



Show solution



  

    
RInteractive R

    
airquality |>
  mutate(was_na    = is.na(Ozone),
         Ozone_imp = if_else(is.na(Ozone), mean(Ozone, na.rm = TRUE), Ozone))

    

      
      
    

    

  

  








Section 4. Outliers (6 problems)



Exercise 4.1: Tukey IQR rule



Difficulty: Intermediate. Flag mpg outliers using Q1 - 1.5IQR / Q3 + 1.5IQR.



Show solution



  

    
RInteractive R

    
mtcars |>
  mutate(out = {
    q <- quantile(mpg, c(0.25, 0.75))
    mpg < q[1] - 1.5*IQR(mpg) | mpg > q[2] + 1.5*IQR(mpg)
  })

    

      
      
    

    

  

  








Exercise 4.2: Z-score rule



Difficulty: Intermediate. Flag rows where |z| > 3 for mpg.



Show solution



  

    
RInteractive R

    
mtcars |>
  mutate(z = (mpg - mean(mpg)) / sd(mpg),
         out = abs(z) > 3)

    

      
      
    

    

  

  








Exercise 4.3: Per-group outliers



Difficulty: Advanced. Flag mpg outliers within each cyl group.



Show solution



  

    
RInteractive R

    
mtcars |>
  group_by(cyl) |>
  mutate(out = {
    q <- quantile(mpg, c(0.25, 0.75))
    mpg < q[1] - 1.5*IQR(mpg) | mpg > q[2] + 1.5*IQR(mpg)
  }) |>
  ungroup()

    

      
      
    

    

  

  








Exercise 4.4: Visualize outliers in a boxplot



Difficulty: Beginner. Boxplot of diamonds$price.



Show solution



  

    
RInteractive R

    
ggplot(diamonds, aes(y = price)) + geom_boxplot()

    

      
      
    

    

  

  








Exercise 4.5: Winsorize



Difficulty: Intermediate. Cap mpg at the 5th and 95th percentiles.



Show solution



  

    
RInteractive R

    
q <- quantile(mtcars$mpg, c(0.05, 0.95))
mtcars |>
  mutate(mpg_w = pmin(pmax(mpg, q[1]), q[2]))

    

      
      
    

    

  

  








Exercise 4.6: Robust scale alternative



Difficulty: Advanced. Standardize using median + MAD instead of mean + sd.



Show solution



  

    
RInteractive R

    
mtcars |>
  mutate(mpg_robust = (mpg - median(mpg)) / mad(mpg))

    

      
      
    

    

  

  








Section 5. Relationships (10 problems)



Exercise 5.1: Pearson correlation



Difficulty: Beginner. Correlation between wt and mpg.



Show solution



  

    
RInteractive R

    
cor(mtcars$wt, mtcars$mpg)

    

      
      
    

    

  

  








Exercise 5.2: Correlation matrix



Difficulty: Intermediate. Correlation matrix of mtcars (numeric).



Show solution



  

    
RInteractive R

    
cor(mtcars)

    

      
      
    

    

  

  








Exercise 5.3: Visualize correlation matrix



Difficulty: Intermediate. Heatmap of the correlation matrix.



Show solution



  

    
RInteractive R

    
cor(mtcars) |>
  as.data.frame() |>
  rownames_to_column("var1") |>
  pivot_longer(-var1, names_to = "var2", values_to = "cor") |>
  ggplot(aes(var1, var2, fill = cor)) +
  geom_tile() +
  scale_fill_gradient2(low = "blue", high = "red", mid = "white", midpoint = 0)

    

      
      
    

    

  

  








Exercise 5.4: Spearman vs Pearson



Difficulty: Intermediate. Compare Pearson and Spearman correlation between disp and mpg.



Show solution



  

    
RInteractive R

    
cor(mtcars$disp, mtcars$mpg, method = "pearson")
cor(mtcars$disp, mtcars$mpg, method = "spearman")
# Spearman captures monotonic non-linear; Pearson assumes linear

    

      
      
    

    

  

  








Exercise 5.5: Scatter with smoother



Difficulty: Intermediate. Scatter wt vs mpg with linear smoother.



Show solution



  

    
RInteractive R

    
ggplot(mtcars, aes(wt, mpg)) +
  geom_point() +
  geom_smooth(method = "lm")

    

      
      
    

    

  

  








Exercise 5.6: Pairs plot



Difficulty: Intermediate. Pairs plot of iris numeric columns colored by Species.



Show solution



  

    
RInteractive R

    
GGally::ggpairs(iris, columns = 1:4, aes(color = Species))

    

      
      
    

    

  

  








Exercise 5.7: Categorical-categorical



Difficulty: Intermediate. Cross-tabulation of cut and clarity in diamonds.



Show solution



  

    
RInteractive R

    
table(diamonds$cut, diamonds$clarity)

    

      
      
    

    

  

  








Exercise 5.8: Categorical-numeric



Difficulty: Intermediate. Mean price per cut (categorical-numeric exploration).



Show solution



  

    
RInteractive R

    
diamonds |>
  group_by(cut) |>
  summarise(mean_price = mean(price))

    

      
      
    

    

  

  








Exercise 5.9: Conditional density



Difficulty: Advanced. Density of mpg conditional on factor(cyl).



Show solution



  

    
RInteractive R

    
ggplot(mtcars, aes(mpg, fill = factor(cyl))) +
  geom_density(alpha = 0.5)

    

      
      
    

    

  

  








Exercise 5.10: Mosaic plot



Difficulty: Advanced. Mosaic plot of cut x clarity proportions.



Show solution



  

    
RInteractive R

    
table(diamonds$cut, diamonds$clarity) |> mosaicplot()

    

      
      
    

    

  

  








Section 6. End-to-end EDA (10 problems)



Exercise 6.1: Initial profile



Difficulty: Intermediate. Run a 3-step opening EDA on diamonds: dim, glimpse, summary.



Show solution



  

    
RInteractive R

    
dim(diamonds); glimpse(diamonds); summary(diamonds)

    

      
      
    

    

  

  








Exercise 6.2: Find a categorical with imbalanced frequencies



Difficulty: Intermediate. Identify any column where the most frequent value is > 50% of rows.



Show solution



  

    
RInteractive R

    
diamonds |>
  summarise(across(c(cut, color, clarity),
                   ~ max(prop.table(table(.x))) > 0.5)) |>
  pivot_longer(everything())

    

      
      
    

    

  

  








Exercise 6.3: Detect a heavily-skewed numeric



Difficulty: Advanced. Find numeric columns with skewness > 1.



Show solution



  

    
RInteractive R

    
diamonds |>
  summarise(across(where(is.numeric), e1071::skewness)) |>
  pivot_longer(everything()) |>
  filter(value > 1)

    

      
      
    

    

  

  








Exercise 6.4: Numeric summary by group



Difficulty: Intermediate. Per Species, give n, mean, sd, min, max of Sepal.Length.



Show solution



  

    
RInteractive R

    
iris |>
  group_by(Species) |>
  summarise(n = n(), mean = mean(Sepal.Length), sd = sd(Sepal.Length),
            min = min(Sepal.Length), max = max(Sepal.Length))

    

      
      
    

    

  

  








Exercise 6.5: Top correlations



Difficulty: Advanced. Find the top 3 most-correlated pairs in mtcars (excluding self).



Show solution



  

    
RInteractive R

    
cor(mtcars) |>
  as.data.frame() |>
  rownames_to_column("var1") |>
  pivot_longer(-var1, names_to = "var2", values_to = "cor") |>
  filter(var1 < var2) |>
  arrange(desc(abs(cor))) |>
  head(3)

    

      
      
    

    

  

  








Exercise 6.6: Detect duplicates



Difficulty: Intermediate. Count fully-duplicate rows in diamonds.



Show solution



  

    
RInteractive R

    
sum(duplicated(diamonds))

    

      
      
    

    

  

  








Exercise 6.7: One-way summary



Difficulty: Intermediate. Mean and N per cyl group with arrange.



Show solution



  

    
RInteractive R

    
mtcars |>
  group_by(cyl) |>
  summarise(n = n(), mean_mpg = mean(mpg)) |>
  arrange(desc(mean_mpg))

    

      
      
    

    

  

  








Exercise 6.8: Two-way summary



Difficulty: Intermediate. Mean price per (cut, color) in diamonds.



Show solution



  

    
RInteractive R

    
diamonds |>
  group_by(cut, color) |>
  summarise(mean_price = mean(price), .groups = "drop")

    

      
      
    

    

  

  








Exercise 6.9: Audit sparse columns



Difficulty: Advanced. List columns where >25% of rows are NA in airquality.



Show solution



  

    
RInteractive R

    
airquality |>
  summarise(across(everything(), ~ mean(is.na(.x)))) |>
  pivot_longer(everything()) |>
  filter(value > 0.25)

    

      
      
    

    

  

  








Exercise 6.10: Decision-quality EDA report



Difficulty: Advanced. Build a one-page EDA: profile + 3 plots (univariate hist, group boxplot, correlation heatmap).



Show solution



  

    
RInteractive R

    
# Profile
print(summary(mtcars))

# Plot 1: distribution
print(ggplot(mtcars, aes(mpg)) + geom_histogram(bins = 15))

# Plot 2: group comparison
print(ggplot(mtcars, aes(factor(cyl), mpg)) + geom_boxplot())

# Plot 3: relationship
print(cor(mtcars) |> as.data.frame() |>
        rownames_to_column("v1") |>
        pivot_longer(-v1, names_to = "v2") |>
        ggplot(aes(v1, v2, fill = value)) + geom_tile() +
        scale_fill_gradient2(low = "blue", high = "red"))

    

      
      
    

    

  

  








What to do next



After 50 EDA problems you should walk into a new dataset and have a profile in 5 minutes. Natural follow-ups:



    


  • Data-Wrangling-Exercises (shipped), the cleaning that EDA reveals.
    • 


  • Linear-Regression-Exercises (shipped), the modeling that EDA precedes.
    • 


  • Data-Visualization-Exercises (coming), viz beyond the EDA basics.
    • 





        


        
      


      

        

        
© 2016-2026 Selva Prabhakaran.
          This work is licensed under the Creative Commons License.
        

        

          About ·
          RSS