Time Complexity of dplyr functions

Thread starter Trollfaz
Start date May 30, 2023

Click For Summary

SUMMARY

The time complexity of basic manipulation functions in R's dplyr package, such as filter(), select(), mutate(), rename(), summarize(), count(), separate(), unite(), spread(), and gather(), is O(N) for a data frame/tibble with n observations. It is established that mutate() operates at O(1). When performing joins between two data frames, df1 and df2, with m rows, the time complexity for inner_join(), left_join(), right_join(), and outer_join() is O(N + M).

PREREQUISITES

Understanding of R programming language
Familiarity with dplyr package functions
Basic knowledge of time complexity analysis
Experience with data frames/tibbles in R

NEXT STEPS

Explore the performance implications of using dplyr's mutate() function
Learn about optimizing joins in dplyr with larger datasets
Investigate the use of dplyr's summarize() for efficient data aggregation
Study advanced dplyr functions such as group_by() and their time complexities

USEFUL FOR

Data analysts, R programmers, and anyone interested in optimizing data manipulation tasks using the dplyr package.

May 30, 2023

Trollfaz

TL;DR: R

For R's dplyr package this is my query.
Suppose I have a data frame/tibble of n observations or n rows. Let's call it df1. Is the time complexity for dplyr's basic manipulation functions O(N)
filter()
select()
mutate() assuming mutate is O(1)
rename()
summarize()
count()
separate()
unite()
spread()
gather()
If I have another data frame/tibble df2 of m rows, then are the following functions of time complexity O(N+M)
inner_join(df1,df2)
right/left_join(df1,df2)
outer_join(df1,df2)