1 Setup

# Install packages 
if (!require("pacman")) install.packages("pacman")

## Loading required package: pacman

pacman::p_load(tidyverse, # tidyverse pkgs including purrr
               tictoc) # performance test 

2 Objective

• Learning how to use reduce() to automate joining multiple dataframes

3 Problem

• How can you make joining multiple dataframes more efficient?
• Note that we will use dplyr::left_join() = merge(x, y, all.x = TRUE).

df1 <- tibble(x = sample(1:10, size = 3, replace = TRUE),
       y = sample(1:10, size = 3, replace = TRUE),
       z = sample(1:10, size = 3, replace = TRUE))

df2 <- tibble(x = sample(1:10, size = 3, replace = TRUE),
       y = sample(1:10, size = 3, replace = TRUE),
       z = sample(1:10, size = 3, replace = TRUE))

df3 <- tibble(x = sample(1:10, size = 3, replace = TRUE),
       y = sample(1:10, size = 3, replace = TRUE),
       z = sample(1:10, size = 3, replace = TRUE))

4 Copy and paste

first_join <- left_join(df1, df2)

## Joining, by = c("x", "y", "z")

second_join <- left_join(first_join, df3)

## Joining, by = c("x", "y", "z")

second_join

Challenge Why the above solution is not efficient?

5 reduce

How reduce() works.

- Input: Takes a vector of length n

- Computation: Calls a function with a pair of values at a time

- Output: Returns a vector of length 1

reduced <- reduce(list(df1, df2, df3), left_join)

## Joining, by = c("x", "y", "z")
## Joining, by = c("x", "y", "z")

reduced