tidyverse in r – Complete Tutorial

tidyverse in R, one of the Important packages in R, there are a lot of new techniques available maybe users are not aware of.

In this tutorial we are importing basic three packages tidyverse, lubridate and nycflights13 for the explanation.

Such tight competition is going around in the data science field, so data analysts should aware of all these kinds of latest techniques.

Load Package

First, we need to load basic three packages into R.


Proportion test in R

Getting Data

Based on nycflights13 data, just load the data into R environment.


tidyverse in r

1. Create a new column basis count option

flights %>%
  mutate(long_flight = (air_time >= 6 * 60)) %>%

You can create new column long flights based on above scripts.

Now need to count the number of long flights

flights %>%
  mutate(long_flight = (air_time >= 6 * 60)) %>%

The above two steps you can execute in a single line.

flights %>%
  count(long_flight = air_time >= 6 * 60)

Same way all different column count can calculate, one example is here.

flights %>%
  count(flight_path = str_c(origin, " -> ", dest), sort = TRUE)

2. Create a new column basis group by

You can create group by summary based on below script.

flights %>%
  group_by(date = make_date(year, month, day)) %>%
  summarise(flights_n = n(), air_time_mean = mean(air_time, na.rm = TRUE)) %>%

3. Randomly Shuffle the data

Suppose you want to randomly slice the data with 15 rows, can execute the same basis below command.

flights %>%
  slice_sample(n = 15)

Using prop command also you can slice the data set.

flights %>%
  slice_sample(prop = 0.15)

4. Date column creation

In the original data set year, month and date contained as separate columns-based make_date command can create new date column.

flights %>%
  select(year, month, day) %>%
  mutate(date = make_date(year, month, day))

5. Number Parsing

Suppose you want extract only numbers then you can you parse_number option.

numbers_1 <- tibble(number = c("#1", "Number8", "How are you 3"))
numbers_1 %>% mutate(number = parse_number(number))

6. Select columns with starts_with and ends_with

You can select the columns based on start_with and end_with option, here is the example

flights %>%
flights %>%
flights %>%

This is one of the useful code for our day to day life.

7. case_when to create when conditions are met

Create a new columns when conditions are met. case_when is one of the handy tool for conditions identification.

one sample analysis in R

flights %>%
  mutate(origin = case_when(
    (origin == "EWR") & dep_delay > 20 ~ "Newark International Airport - DELAYED",
    (origin == "EWR") & dep_delay <= 20 ~ "Newark International Airport - ON TIME DEPARTURE",
  )) %>%

8. str_replace_all to find and replace multiple options at once

Every one aware about str_replace in string r pacakage, here we can execute replace multiple options at a once.

flights %>%
  mutate(origin = str_replace_all(origin, c(
    "^EWR$" = "Newark International",    "^JFK$" = "John F. Kennedy International"
  ))) %>%

9. Filter groups without making a new column

Filtering is one of the essential function for cleaning and checking data sets.

flights_top_carriers <- flights %>%
  group_by(carrier) %>%
  filter(n() >= 10000) %>%

10. Extract rows from the first table which are matched in the second table

You can extract the row information’s based on str_detect function

beginning_with_am<- airlines %>%   
filter(name %>% str_detect("^Am")) 

Non Parametric tests

11. Extract rows from the first table which are not matched in the second table

Same way you can remove row information’s from the data frame while using anti_join function

flights %>%
  anti_join(airways_beginning_with_a, by = "carrier")

12. fct_reorder to sort for charts creation

When you are creating graphs reordering one of the key function, tidyverse will handle such kind of situations.

airline_names <- flights %>%
  left_join(airlines, by = "carrier")
airline_names %>%
  count(name) %>%
  ggplot(aes(name, n)) +
airline_names %>%
  count(name) %>%
  mutate(name = fct_reorder(name, n)) %>%
  ggplot(aes(name, n)) +

 13. coord_flip to display counts more accurately

To change x and y axis and make a beautiful display

flights_with_airline_names %>%   
count(name) %>%   
mutate(name = fct_reorder(name, n)) %>%   
ggplot(aes(name, n)) +   
geom_col() +   

Types of Data Visualization

14. Generate all combinations using crossing

Like expand grid in R, you can create all possible combinations based on crossing function in tidyverse.

  customer_channel = c("Bus", "Car"),
  customer_status = c("New", "Repeat"),
  spend_range = c("$0-$10", "$10-$20", "$20-$50", "$50+"))

15. Group by based on function

Write the function based on your requirements and group by accordingly.

summary <- function(data, col_names, na.rm = TRUE) {
  data %>%
    summarise(across({{ col_names }},
                       min = min,
                       max = max,
                       median = median,
                       mean = mean
                     na.rm = na.rm,
                     .names = "{col}_{fn}"
flights_with_airline_names %>%
  summary(c(air_time, arr_delay))
flights_with_airline_names %>%
  group_by(carrier) %>%
summary(c(air_time, arr_delay))

Uses of Index Numbers

About the author

Data Specialist

Leave a Reply

Your email address will not be published.

twenty + 19 =