dplyr presentation at Budapest BI Forum 2015.

1. dplyr
@romain_francois

2. • Use R since 2002
• #rcatladies
• R Enthusiast
• R/C++ hero
• Performance
• dplyr
• Occasional comedy

5. %>%from magrittr

6. enjoy(cool(bake(shape(beat(append(bowl(rep("flour",
2), "yeast", "water", "milk", "oil"), "flour", until
= "soft"), duration = "3mins"), as = "balls", style =
"slightly-flat"), degrees = 200, duration =
"15mins"), duration = "5mins"))
bowl(rep("flour", 2), "yeast", "water", "milk", "oil") %>%
append("flour", until = "soft") %>%
beat(duration = "3mins") %>%
shape(as = "balls", style = "slightly-flat") %>%
bake(degrees = 200, duration = "15mins") %>%
cool(buns, duration = "5mins") %>%
enjoy()

7. nycflights13
> flights
Source: local data frame [336,776 x 16]
year month day dep_time dep_delay arr_time arr_delay carrier tailnum flight
1 2013 1 1 517 2 830 11 UA N14228 1545
2 2013 1 1 533 4 850 20 UA N24211 1714
.. ... ... ... ... ... ... ... ... ... ...
Variables not shown: origin (chr), dest (chr), air_time (dbl), distance (dbl),
hour (dbl), minute (dbl)

8. nycflights13
> glimpse(flights)
Observations: 336,776
Variables: 16
$ year (int) 2013, 2013, 2013, 2013, 2013, 2013, 2013, 2013, 2013, 201...
$ month (int) 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, ...
$ day (int) 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, ...
$ dep_time (int) 517, 533, 542, 544, 554, 554, 555, 557, 557, 558, 558, 55...
$ dep_delay (dbl) 2, 4, 2, -1, -6, -4, -5, -3, -3, -2, -2, -2, -2, -2, -1, ...
$ arr_time (int) 830, 850, 923, 1004, 812, 740, 913, 709, 838, 753, 849, 8...
$ arr_delay (dbl) 11, 20, 33, -18, -25, 12, 19, -14, -8, 8, -2, -3, 7, -14,...
$ carrier (chr) "UA", "UA", "AA", "B6", "DL", "UA", "B6", "EV", "B6", "AA...
$ tailnum (chr) "N14228", "N24211", "N619AA", "N804JB", "N668DN", "N39463...
$ flight (int) 1545, 1714, 1141, 725, 461, 1696, 507, 5708, 79, 301, 49,...
$ origin (chr) "EWR", "LGA", "JFK", "JFK", "LGA", "EWR", "EWR", "LGA", "...
$ dest (chr) "IAH", "IAH", "MIA", "BQN", "ATL", "ORD", "FLL", "IAD", "...
$ air_time (dbl) 227, 227, 160, 183, 116, 150, 158, 53, 140, 138, 149, 158...
$ distance (dbl) 1400, 1416, 1089, 1576, 762, 719, 1065, 229, 944, 733, 10...
$ hour (dbl) 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 6, 6, ...
$ minute (dbl) 17, 33, 42, 44, 54, 54, 55, 57, 57, 58, 58, 58, 58, 58, 5...

9. filterA subset of the rows of the data frame
flights %>%
filter( dep_delay < 10 )
flights %>%
filter( arr_delay < dep_delay )

10. slicefilter rows by position
flights %>%
slice( 1:10 )

11. arrangereorder a data frame
flights %>%
filter( hour < 8 ) %>%
arrange( year, month, day )

12. selectselect certain columns from the data frame
select(flights, year, month, day)
select(flights, year:day)
select(flights, -(year:day))

13. mutatemodify or create columns based on others
flights %>%
mutate(
gain = arr_delay - dep_delay,
speed = distance / air_time * 60
) %>%
filter( gain > 0 ) %>%
arrange( desc(speed) ) %>%
select( year, month, day, dest, gain, speed )

14. summarisecollapse a data frame into one row …
flights %>%
summarise(delay = mean(dep_delay, na.rm = TRUE))
flights %>%
filter( dep_delay > 0 ) %>%
summarise(arr_delay = mean(arr_delay, na.rm = TRUE))

15. group_byGroup observations by one or more variables
flights %>%
group_by( tailnum ) %>%
summarise(
count = n(),
dist = mean(distance, na.rm = TRUE),
delay = mean(arr_delay, na.rm = TRUE)
) %>%
filter( is.finite(delay) ) %>%
arrange( desc(count) )

16. bind_rows
bind_rows( , )
color num
green 1
yellow 2
red 3
blue 4
pink 5
color num
green 1
yellow 2
color num
red 3
blue 4
pink 5

17. joins
a <- data_frame(
color = c("green", "yellow", "red"),
num = 1:3
)
b <- data_frame(
color = c("green", "yellow", "pink"),
size = c("S", "M", "L")
)
color num
green 1
yellow 2
red 3
color size
green S
yellow M
pink L

18. inner_join
color num
green 1
yellow 2
red 3
color size
green S
yellow M
pink L
inner_join( , )
color num size
green 1 S
yellow 2 M

19. left_join
color num
green 1
yellow 2
red 3
color size
green S
yellow M
pink L
left_join( , )
color num size
green 1 S
yellow 2 M
red 3

20. right_join
color num
green 1
yellow 2
red 3
color size
green S
yellow M
pink L
right_join( , )
color num size
green 1 S
yellow 2 M
pink L

21. full_join
color num
green 1
yellow 2
red 3
color size
green S
yellow M
pink L
full_join( , )
color num size
green 1 S
yellow 2 M
red 3
pink L

22. data_frameJust like data.frame, but better
> data_frame( x = 1:5, y = letters[1:5] ) %>% glimpse
Observations: 5
Variables: 2
$ x (int) 1, 2, 3, 4, 5
$ y (chr) "a", "b", "c", "d", "e"
> data_frame( x = 1:5, y = letters[1:5] , z = x + 1) %>% glimpse
Observations: 5
Variables: 3
$ x (int) 1, 2, 3, 4, 5
$ y (chr) "a", "b", "c", "d", "e"
$ z (dbl) 2, 3, 4, 5, 6

23. frame_data aka tibble
> frame_data(
+ ~colA, ~colB,
+ "a", 1,
+ "b", 2
+ )
Source: local data frame [2 x 2]
colA colB
(chr) (dbl)
1 a 1
2 b 2

24. _

25. g <- c("origin", "dest")
v <- "dep_delay"
flights %>%
group_by( g ) %>%
summarise( result = mean(v, na.rm = TRUE) )
🙀
🙀

26. g <- c("origin", "dest")
v <- "dep_delay"
flights %>%
group_by_( .dots = g ) %>%
summarise_( .dots =
interp(~ mean(var, na.rm = TRUE), var = as.name(v))
)

30. Future
• Performance improvements (parallel C++)
• Alternative back ends
• Different type of groupings (e.g. bootstrap)
As soon as we get hoverboard ...

31. dplyr
Romain François
@romain_francois
romain@r-enthusiasts.com