All notes

5.1 Introduction

• filter, pick observations
• arrange, reorder the row
• select, pick the variables by name
• mutate, create variable with functions of existing variables
• summarise, Collapse many values down to a single summary
• group_by, used in conjunction with the above, apply it on different groups

Use them similarly:

1. The first arg is a dataframe
2. The subsequent arguments describe what to do with the DF, variable names without quote
3. result is a new DF.

5.2 filter() notes

• If we want to both assign to a variable and print out, use ()
• Comparison operators: >, >=, <, <=, !=, ==
  • when compare numerical, use near(sqrt(2)^2, 2)
• Logic operators: &, |, !
  • handy shorthand: x %in% y
• Almost all operations with NA is NA.
• Filter only includes rows where the condition is TRUE, excludes both NA and FALSE

5.3 arrange() notes

• arrange() change the order of the rows.
• use desc() to reverse the order.
• NA is always at the end, no matter ascending or desending

5.4 select() notes

• Select couple of variables
• Have a couple of helper functions
  • Like starts_with and matches, see ?select_helpers
• rename function to change variable name
• everthing helper function to bring a variable to the front

5.5 mutate() notes

• mutate() can use the variable just created
• transmute() only keeps the new variables
• Useful creation functions, take in vectors and produce vectors
  • Arithmetic operators: +/-/*/ / /^
  • Arithmetic operators with aggregate functions: x/sum(x)
  • module %/%, %%
  • logs
  • offset: lead/lag, useful in computing running difference or find when value changes.
  • cumulative function
  • ranking functions: min_rank() and etc. see ?min_rank

5.6 summarise() notes

• collapse a data frame to a single row
• More useful with group_by
• Use pipe %>% is much easier to coding and reading
  • x %>% f(y) %>% g(z) <=> g(f(x, y), z)
  • example here:

delays <- flights %>% 
  group_by(dest) %>% 
  summarise(
    count = n(),
    dist = mean(distance, na.rm = TRUE),
    delay = mean(arr_delay, na.rm = TRUE)
  ) %>% 
  filter(count > 20, dest != "HNL")

• missing value, mean function has na.rm.
  • But you can also do filter like

5.2 filter()

5.2.1 Exercise