week2_codingPractice2
Leona Chase
2024-09-07
Week 2 Coding Practice R, Part 2: Data Transformation
5.1 Introduction
nycflights13
Dataframe contains all 336,776 flights that departed from NYC in 2013, data provided by the US Bureau of Transportation. DataFrames printed as a tibble which are dataframes but tweaked to work better in the tidyverse.
flights## # A tibble: 336,776 × 19
## year month day dep_time sched_dep_time dep_delay arr_time sched_arr_time
## <int> <int> <int> <int> <int> <dbl> <int> <int>
## 1 2013 1 1 517 515 2 830 819
## 2 2013 1 1 533 529 4 850 830
## 3 2013 1 1 542 540 2 923 850
## 4 2013 1 1 544 545 -1 1004 1022
## 5 2013 1 1 554 600 -6 812 837
## 6 2013 1 1 554 558 -4 740 728
## 7 2013 1 1 555 600 -5 913 854
## 8 2013 1 1 557 600 -3 709 723
## 9 2013 1 1 557 600 -3 838 846
## 10 2013 1 1 558 600 -2 753 745
## # ℹ 336,766 more rows
## # ℹ 11 more variables: arr_delay <dbl>, carrier <chr>, flight <int>,
## # tailnum <chr>, origin <chr>, dest <chr>, air_time <dbl>, distance <dbl>,
## # hour <dbl>, minute <dbl>, time_hour <dttm>dplyr basics
- filter(): Pick observations based on their values
- arrange(): Reorder the rows
- select(): Pick variables by their names
- mutate(): Create new variables with functions of existing variables
- summarise(): Collapse many values down to a single summary
These can be used in conjunction with group_by()
which changes the scope of each function from operating on the entire
dataset by operating on it group-by-group.
5.2 Using filter()
Select all flights on January 1st:
filter(flights, month == 1, day == 1)## # A tibble: 842 × 19
## year month day dep_time sched_dep_time dep_delay arr_time sched_arr_time
## <int> <int> <int> <int> <int> <dbl> <int> <int>
## 1 2013 1 1 517 515 2 830 819
## 2 2013 1 1 533 529 4 850 830
## 3 2013 1 1 542 540 2 923 850
## 4 2013 1 1 544 545 -1 1004 1022
## 5 2013 1 1 554 600 -6 812 837
## 6 2013 1 1 554 558 -4 740 728
## 7 2013 1 1 555 600 -5 913 854
## 8 2013 1 1 557 600 -3 709 723
## 9 2013 1 1 557 600 -3 838 846
## 10 2013 1 1 558 600 -2 753 745
## # ℹ 832 more rows
## # ℹ 11 more variables: arr_delay <dbl>, carrier <chr>, flight <int>,
## # tailnum <chr>, origin <chr>, dest <chr>, air_time <dbl>, distance <dbl>,
## # hour <dbl>, minute <dbl>, time_hour <dttm>To save the results:
jan1 <- filter(flights, month == 1, day == 1)To print AND save them to a variable together, wrap in parentheses:
(dec25 <- filter(flights, month == 12, day == 25))## # A tibble: 719 × 19
## year month day dep_time sched_dep_time dep_delay arr_time sched_arr_time
## <int> <int> <int> <int> <int> <dbl> <int> <int>
## 1 2013 12 25 456 500 -4 649 651
## 2 2013 12 25 524 515 9 805 814
## 3 2013 12 25 542 540 2 832 850
## 4 2013 12 25 546 550 -4 1022 1027
## 5 2013 12 25 556 600 -4 730 745
## 6 2013 12 25 557 600 -3 743 752
## 7 2013 12 25 557 600 -3 818 831
## 8 2013 12 25 559 600 -1 855 856
## 9 2013 12 25 559 600 -1 849 855
## 10 2013 12 25 600 600 0 850 846
## # ℹ 709 more rows
## # ℹ 11 more variables: arr_delay <dbl>, carrier <chr>, flight <int>,
## # tailnum <chr>, origin <chr>, dest <chr>, air_time <dbl>, distance <dbl>,
## # hour <dbl>, minute <dbl>, time_hour <dttm>Comparisons
R uses the standard suite of comparison operators: * > * >= * < * <= * != (not equal) * == (equal)
Basic Beginner Error: = does not mean ==
# WRONG:
# filter(flights, month = 1)
# CORRECT:
filter(flights, month == 1)## # A tibble: 27,004 × 19
## year month day dep_time sched_dep_time dep_delay arr_time sched_arr_time
## <int> <int> <int> <int> <int> <dbl> <int> <int>
## 1 2013 1 1 517 515 2 830 819
## 2 2013 1 1 533 529 4 850 830
## 3 2013 1 1 542 540 2 923 850
## 4 2013 1 1 544 545 -1 1004 1022
## 5 2013 1 1 554 600 -6 812 837
## 6 2013 1 1 554 558 -4 740 728
## 7 2013 1 1 555 600 -5 913 854
## 8 2013 1 1 557 600 -3 709 723
## 9 2013 1 1 557 600 -3 838 846
## 10 2013 1 1 558 600 -2 753 745
## # ℹ 26,994 more rows
## # ℹ 11 more variables: arr_delay <dbl>, carrier <chr>, flight <int>,
## # tailnum <chr>, origin <chr>, dest <chr>, air_time <dbl>, distance <dbl>,
## # hour <dbl>, minute <dbl>, time_hour <dttm>== can cause problems with floating points.
Computers use finite precision arithmetic, cannot store infinite digits. Each number is an approx so instead of ==, use near()
sqrt(2) ^ 2 == 2## [1] FALSE1 / 49 * 49 == 1## [1] FALSEnear(sqrt(2) ^ 2, 2)## [1] TRUEnear(1 / 49 * 49, 1)## [1] TRUELogical Operators
Combine multiple arguments in filter() with logical operators:
- & = “and”
- | = “or”
- ! = “not”
Find all flights that departed in November OR December:
# Incorrect:
# filter(flights, month == (11 | 12))
# Correct:
filter(flights, month == 11 | month == 12)## # A tibble: 55,403 × 19
## year month day dep_time sched_dep_time dep_delay arr_time sched_arr_time
## <int> <int> <int> <int> <int> <dbl> <int> <int>
## 1 2013 11 1 5 2359 6 352 345
## 2 2013 11 1 35 2250 105 123 2356
## 3 2013 11 1 455 500 -5 641 651
## 4 2013 11 1 539 545 -6 856 827
## 5 2013 11 1 542 545 -3 831 855
## 6 2013 11 1 549 600 -11 912 923
## 7 2013 11 1 550 600 -10 705 659
## 8 2013 11 1 554 600 -6 659 701
## 9 2013 11 1 554 600 -6 826 827
## 10 2013 11 1 554 600 -6 749 751
## # ℹ 55,393 more rows
## # ℹ 11 more variables: arr_delay <dbl>, carrier <chr>, flight <int>,
## # tailnum <chr>, origin <chr>, dest <chr>, air_time <dbl>, distance <dbl>,
## # hour <dbl>, minute <dbl>, time_hour <dttm># Another Correct option:
nov_dec <- filter(flights, month %in% c(11, 12))Find flights that weren’t delated on arrival or departure) by more than 2 hours
# Correct:
filter(flights, !(arr_delay > 120 | dep_delay > 120))## # A tibble: 316,050 × 19
## year month day dep_time sched_dep_time dep_delay arr_time sched_arr_time
## <int> <int> <int> <int> <int> <dbl> <int> <int>
## 1 2013 1 1 517 515 2 830 819
## 2 2013 1 1 533 529 4 850 830
## 3 2013 1 1 542 540 2 923 850
## 4 2013 1 1 544 545 -1 1004 1022
## 5 2013 1 1 554 600 -6 812 837
## 6 2013 1 1 554 558 -4 740 728
## 7 2013 1 1 555 600 -5 913 854
## 8 2013 1 1 557 600 -3 709 723
## 9 2013 1 1 557 600 -3 838 846
## 10 2013 1 1 558 600 -2 753 745
## # ℹ 316,040 more rows
## # ℹ 11 more variables: arr_delay <dbl>, carrier <chr>, flight <int>,
## # tailnum <chr>, origin <chr>, dest <chr>, air_time <dbl>, distance <dbl>,
## # hour <dbl>, minute <dbl>, time_hour <dttm># Another Correct option:
filter(flights, arr_delay <= 120, dep_delay <= 120)## # A tibble: 316,050 × 19
## year month day dep_time sched_dep_time dep_delay arr_time sched_arr_time
## <int> <int> <int> <int> <int> <dbl> <int> <int>
## 1 2013 1 1 517 515 2 830 819
## 2 2013 1 1 533 529 4 850 830
## 3 2013 1 1 542 540 2 923 850
## 4 2013 1 1 544 545 -1 1004 1022
## 5 2013 1 1 554 600 -6 812 837
## 6 2013 1 1 554 558 -4 740 728
## 7 2013 1 1 555 600 -5 913 854
## 8 2013 1 1 557 600 -3 709 723
## 9 2013 1 1 557 600 -3 838 846
## 10 2013 1 1 558 600 -2 753 745
## # ℹ 316,040 more rows
## # ℹ 11 more variables: arr_delay <dbl>, carrier <chr>, flight <int>,
## # tailnum <chr>, origin <chr>, dest <chr>, air_time <dbl>, distance <dbl>,
## # hour <dbl>, minute <dbl>, time_hour <dttm>Missing Values
NAs = “Not Availables”. Represent an unknown value and are “contagious”. Almost any operation involving an uknown value will also be unknown.
NA > 5## [1] NA10 == NA## [1] NANA + 10## [1] NANA / 2## [1] NANA == NA## [1] NAContextually:
# Let x be Mary's age. We don't know how old she is.
x <- NA
# Let y be John's age. We don't know how old he is.
y <- NA
# Are John and Mary the same age? We don't know!
x == y## [1] NADetermine if a value is missing:
is.na(x)## [1] TRUEFilter() excludes both FALSE and NA values. To preserve missing values, ask explicitly:
# Create df
df <- tibble(x = c(1, NA, 3))
df ## # A tibble: 3 × 1
## x
## <dbl>
## 1 1
## 2 NA
## 3 3# filter df to show only values > 1
filter(df, x > 1)## # A tibble: 1 × 1
## x
## <dbl>
## 1 3# filter df to show only values that are NA or > 1
filter(df, is.na(x) | x > 1)## # A tibble: 2 × 1
## x
## <dbl>
## 1 NA
## 2 35.3 Arrange Rows with arrange()
arrange() works similarly to filter() except instead of selecting rows, it changes their order. Takes a dataframe and a set of column names (or more complicated expressions) to order by. If more than 1 column name is provided, each additional column will act as a tie-breaker in values of preceding columns.
Arrange flights by year, month, then day
arrange(flights, year, month, day)## # A tibble: 336,776 × 19
## year month day dep_time sched_dep_time dep_delay arr_time sched_arr_time
## <int> <int> <int> <int> <int> <dbl> <int> <int>
## 1 2013 1 1 517 515 2 830 819
## 2 2013 1 1 533 529 4 850 830
## 3 2013 1 1 542 540 2 923 850
## 4 2013 1 1 544 545 -1 1004 1022
## 5 2013 1 1 554 600 -6 812 837
## 6 2013 1 1 554 558 -4 740 728
## 7 2013 1 1 555 600 -5 913 854
## 8 2013 1 1 557 600 -3 709 723
## 9 2013 1 1 557 600 -3 838 846
## 10 2013 1 1 558 600 -2 753 745
## # ℹ 336,766 more rows
## # ℹ 11 more variables: arr_delay <dbl>, carrier <chr>, flight <int>,
## # tailnum <chr>, origin <chr>, dest <chr>, air_time <dbl>, distance <dbl>,
## # hour <dbl>, minute <dbl>, time_hour <dttm>Arrange dep_delay in descending order
arrange(flights, desc(dep_delay))## # A tibble: 336,776 × 19
## year month day dep_time sched_dep_time dep_delay arr_time sched_arr_time
## <int> <int> <int> <int> <int> <dbl> <int> <int>
## 1 2013 1 9 641 900 1301 1242 1530
## 2 2013 6 15 1432 1935 1137 1607 2120
## 3 2013 1 10 1121 1635 1126 1239 1810
## 4 2013 9 20 1139 1845 1014 1457 2210
## 5 2013 7 22 845 1600 1005 1044 1815
## 6 2013 4 10 1100 1900 960 1342 2211
## 7 2013 3 17 2321 810 911 135 1020
## 8 2013 6 27 959 1900 899 1236 2226
## 9 2013 7 22 2257 759 898 121 1026
## 10 2013 12 5 756 1700 896 1058 2020
## # ℹ 336,766 more rows
## # ℹ 11 more variables: arr_delay <dbl>, carrier <chr>, flight <int>,
## # tailnum <chr>, origin <chr>, dest <chr>, air_time <dbl>, distance <dbl>,
## # hour <dbl>, minute <dbl>, time_hour <dttm>Missing Values are always sorted at the end, regardless of asc, desc
df <- tibble(x = c(5, 2, NA))
arrange(df,x)## # A tibble: 3 × 1
## x
## <dbl>
## 1 2
## 2 5
## 3 NAarrange(df, desc(x))## # A tibble: 3 × 1
## x
## <dbl>
## 1 5
## 2 2
## 3 NA5.4 Select Columns with select()
select() allows you to rapidly zoom in on a useful subset using operations based on the names of the variables.
Select columns by name
select(flights, year, month, day)## # A tibble: 336,776 × 3
## year month day
## <int> <int> <int>
## 1 2013 1 1
## 2 2013 1 1
## 3 2013 1 1
## 4 2013 1 1
## 5 2013 1 1
## 6 2013 1 1
## 7 2013 1 1
## 8 2013 1 1
## 9 2013 1 1
## 10 2013 1 1
## # ℹ 336,766 more rowsSelect all columns between year and day (inclusive)
select(flights, year:day)## # A tibble: 336,776 × 3
## year month day
## <int> <int> <int>
## 1 2013 1 1
## 2 2013 1 1
## 3 2013 1 1
## 4 2013 1 1
## 5 2013 1 1
## 6 2013 1 1
## 7 2013 1 1
## 8 2013 1 1
## 9 2013 1 1
## 10 2013 1 1
## # ℹ 336,766 more rowsSelect all columns except those from year and day (inclusive)
select(flights, -(year:day))## # A tibble: 336,776 × 16
## dep_time sched_dep_time dep_delay arr_time sched_arr_time arr_delay carrier
## <int> <int> <dbl> <int> <int> <dbl> <chr>
## 1 517 515 2 830 819 11 UA
## 2 533 529 4 850 830 20 UA
## 3 542 540 2 923 850 33 AA
## 4 544 545 -1 1004 1022 -18 B6
## 5 554 600 -6 812 837 -25 DL
## 6 554 558 -4 740 728 12 UA
## 7 555 600 -5 913 854 19 B6
## 8 557 600 -3 709 723 -14 EV
## 9 557 600 -3 838 846 -8 B6
## 10 558 600 -2 753 745 8 AA
## # ℹ 336,766 more rows
## # ℹ 9 more variables: flight <int>, tailnum <chr>, origin <chr>, dest <chr>,
## # air_time <dbl>, distance <dbl>, hour <dbl>, minute <dbl>, time_hour <dttm>Helper Functions with select()
- starts_with(“abc”): matches names that begin with “abc”.
- ends_with(“xyz”): matches names that end with “xyz”.
- contains(“ijk”): matches names that contain “ijk”.
- matches(“(.)\1”): selects variables that match a regular expression. This one matches any variables that contain repeated characters.
- num_range(“x”, 1:3): matches x1, x2 and x3.
Note: select() can be used to rename variables, but it’s rarely useful because it drops all of the variables not explicitly mentioned. Instead, use rename(), which is a variant of select() that keeps all the variables that aren’t explicitly mentioned:
select() in conjunction with everything to move some variables to the start of the df.
select(flights, time_hour, air_time, everything())## # A tibble: 336,776 × 19
## time_hour air_time year month day dep_time sched_dep_time
## <dttm> <dbl> <int> <int> <int> <int> <int>
## 1 2013-01-01 05:00:00 227 2013 1 1 517 515
## 2 2013-01-01 05:00:00 227 2013 1 1 533 529
## 3 2013-01-01 05:00:00 160 2013 1 1 542 540
## 4 2013-01-01 05:00:00 183 2013 1 1 544 545
## 5 2013-01-01 06:00:00 116 2013 1 1 554 600
## 6 2013-01-01 05:00:00 150 2013 1 1 554 558
## 7 2013-01-01 06:00:00 158 2013 1 1 555 600
## 8 2013-01-01 06:00:00 53 2013 1 1 557 600
## 9 2013-01-01 06:00:00 140 2013 1 1 557 600
## 10 2013-01-01 06:00:00 138 2013 1 1 558 600
## # ℹ 336,766 more rows
## # ℹ 12 more variables: dep_delay <dbl>, arr_time <int>, sched_arr_time <int>,
## # arr_delay <dbl>, carrier <chr>, flight <int>, tailnum <chr>, origin <chr>,
## # dest <chr>, distance <dbl>, hour <dbl>, minute <dbl>5.5 Add new variables with mutate()
mutate() always adds columns at the end of your dataset. Note: In RStudio, the easiest way to see all columns is View().
Create a smaller df so it’s easier to see new column at the end:
flights_sml <- select(flights,
year:day,
ends_with("delay"),
distance,
air_time
)Calculate new columns gain and speed:
mutate(flights_sml,
gain = dep_delay - arr_delay,
speed = distance / air_time * 60
)## # A tibble: 336,776 × 9
## year month day dep_delay arr_delay distance air_time gain speed
## <int> <int> <int> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl>
## 1 2013 1 1 2 11 1400 227 -9 370.
## 2 2013 1 1 4 20 1416 227 -16 374.
## 3 2013 1 1 2 33 1089 160 -31 408.
## 4 2013 1 1 -1 -18 1576 183 17 517.
## 5 2013 1 1 -6 -25 762 116 19 394.
## 6 2013 1 1 -4 12 719 150 -16 288.
## 7 2013 1 1 -5 19 1065 158 -24 404.
## 8 2013 1 1 -3 -14 229 53 11 259.
## 9 2013 1 1 -3 -8 944 140 5 405.
## 10 2013 1 1 -2 8 733 138 -10 319.
## # ℹ 336,766 more rowsYou can refer to the columns created by mutate():
mutate(flights_sml,
gain = dep_delay - arr_delay,
hours = air_time / 60,
gain_per_hour = gain / hours
)## # A tibble: 336,776 × 10
## year month day dep_delay arr_delay distance air_time gain hours
## <int> <int> <int> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl>
## 1 2013 1 1 2 11 1400 227 -9 3.78
## 2 2013 1 1 4 20 1416 227 -16 3.78
## 3 2013 1 1 2 33 1089 160 -31 2.67
## 4 2013 1 1 -1 -18 1576 183 17 3.05
## 5 2013 1 1 -6 -25 762 116 19 1.93
## 6 2013 1 1 -4 12 719 150 -16 2.5
## 7 2013 1 1 -5 19 1065 158 -24 2.63
## 8 2013 1 1 -3 -14 229 53 11 0.883
## 9 2013 1 1 -3 -8 944 140 5 2.33
## 10 2013 1 1 -2 8 733 138 -10 2.3
## # ℹ 336,766 more rows
## # ℹ 1 more variable: gain_per_hour <dbl>To keep only the new variables, use transmute():
transmute(flights,
gain = dep_delay - arr_delay,
hours = air_time / 60,
gain_per_hour = gain / hours
)## # A tibble: 336,776 × 3
## gain hours gain_per_hour
## <dbl> <dbl> <dbl>
## 1 -9 3.78 -2.38
## 2 -16 3.78 -4.23
## 3 -31 2.67 -11.6
## 4 17 3.05 5.57
## 5 19 1.93 9.83
## 6 -16 2.5 -6.4
## 7 -24 2.63 -9.11
## 8 11 0.883 12.5
## 9 5 2.33 2.14
## 10 -10 2.3 -4.35
## # ℹ 336,766 more rowsUseful Creation Functions
Key property is that the function must be vectorised: it must take a vector of values as input, return a vector with the same number of values as output.
Arithmetic Operators: Especially useful in conjunction with aggregate functions like mean
- +
- -
- *
- /
- ^
Modular Arithmetic: Allows you to break integers into pieces
- %/% - Integer Division
- %% - Remainder where x == y * (x %/% y) + (x %% y)
transmute(flights,
dep_time,
hour = dep_time %/% 100,
minute = dep_time %% 100
)## # A tibble: 336,776 × 3
## dep_time hour minute
## <int> <dbl> <dbl>
## 1 517 5 17
## 2 533 5 33
## 3 542 5 42
## 4 544 5 44
## 5 554 5 54
## 6 554 5 54
## 7 555 5 55
## 8 557 5 57
## 9 557 5 57
## 10 558 5 58
## # ℹ 336,766 more rowsLogs: Useful for dealing wiht data that ranges across multiple order of magnitude.
- log()
- log2()
- log10()
Offsets: Allow you to refer to leading or lagging values, often used with group_by()
- lead()
- lag()
(x <- 1:10)## [1] 1 2 3 4 5 6 7 8 9 10lag(x)## [1] NA 1 2 3 4 5 6 7 8 9lead(x)## [1] 2 3 4 5 6 7 8 9 10 NACumulative and Rolling Aggregates: Runing sums, products, mins, maxes
- cumsum()
- cumprod()
- cummin()
- cummax()
- cummean() - dplyr package
x## [1] 1 2 3 4 5 6 7 8 9 10cumsum(x)## [1] 1 3 6 10 15 21 28 36 45 55cummean(x)## [1] 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5Ranking
- min_rank()
- row_number()
- dense_rank()
- percent_rank()
- cume_dist()
- ntile()
y <- c(1, 2, 2, NA, 3, 4)
min_rank(y)## [1] 1 2 2 NA 4 5min_rank(desc(y))## [1] 5 3 3 NA 2 1row_number(y)## [1] 1 2 3 NA 4 5dense_rank(y)## [1] 1 2 2 NA 3 4percent_rank(y)## [1] 0.00 0.25 0.25 NA 0.75 1.00cume_dist(y)## [1] 0.2 0.6 0.6 NA 0.8 1.05.6 Grouped summaries with summarise()
Collapses a data frame to a single row. More useful when paired with group_by() This changes the unit of analysis form the complete dataset to individual groups.
Summarize Mean of dep_delay for the whole dataset
summarise(flights, delay = mean(dep_delay, na.rm = TRUE))## # A tibble: 1 × 1
## delay
## <dbl>
## 1 12.6Summarize but group by data for the average delay per date
by_day <- group_by(flights, year, month, day)
summarise(by_day, delay = mean(dep_delay, na.rm = TRUE))## `summarise()` has grouped output by 'year', 'month'. You can override using the
## `.groups` argument.## # A tibble: 365 × 4
## # Groups: year, month [12]
## year month day delay
## <int> <int> <int> <dbl>
## 1 2013 1 1 11.5
## 2 2013 1 2 13.9
## 3 2013 1 3 11.0
## 4 2013 1 4 8.95
## 5 2013 1 5 5.73
## 6 2013 1 6 7.15
## 7 2013 1 7 5.42
## 8 2013 1 8 2.55
## 9 2013 1 9 2.28
## 10 2013 1 10 2.84
## # ℹ 355 more rowsCombining Multiple operations with the Pipe
Say we want to explore the relationship between the distance and average delay for each location.
We could do prepare the data in 3 steps:
- Group flights by destination.
- Summarise to ompute distance, average delay, and number of flights
- Filter to remove noisy points and Honolulu airport, which is almost twice as far away as the next closest airport.
This process involves the extra work of creating and naming variables that we really don’t care about, slowing the process down.
by_dest <- group_by(flights, dest)
delay <- summarise(by_dest,
count = n(),
dist = mean(distance, na.rm = TRUE),
delay = mean(arr_delay, na.rm = TRUE)
)
delay <- filter(delay, count > 20, dest != "HNL")
# It looks like delays increase with distance up to ~750 miles
# and then decrease. Maybe as flights get longer there's more
# ability to make up delays in the air?
ggplot(data = delay, mapping = aes(x = dist, y = delay)) +
geom_point(aes(size = count), alpha = 1/3) +
geom_smooth(se = FALSE)## `geom_smooth()` using method = 'loess' and formula = 'y ~ x'
Instead, we could use the pipe, %>%, to focus on the transformation, not what’s being transformed, making the code easier to read.
Pipe works in a series of impertaive statements: group, then summarise, then filter.
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 values
Aggregation functions obey the usual rule of missing values: if there are any missing values in the input, the output will be a missing value.
flights %>%
group_by(year, month, day) %>%
summarise(mean = mean(dep_delay))## `summarise()` has grouped output by 'year', 'month'. You can override using the
## `.groups` argument.## # A tibble: 365 × 4
## # Groups: year, month [12]
## year month day mean
## <int> <int> <int> <dbl>
## 1 2013 1 1 NA
## 2 2013 1 2 NA
## 3 2013 1 3 NA
## 4 2013 1 4 NA
## 5 2013 1 5 NA
## 6 2013 1 6 NA
## 7 2013 1 7 NA
## 8 2013 1 8 NA
## 9 2013 1 9 NA
## 10 2013 1 10 NA
## # ℹ 355 more rowsTo fix this, use na.rm=TRUE argument to remove the missing values prior to computations:
flights %>%
group_by(year, month, day) %>%
summarise(mean = mean(dep_delay, na.rm = TRUE))## `summarise()` has grouped output by 'year', 'month'. You can override using the
## `.groups` argument.## # A tibble: 365 × 4
## # Groups: year, month [12]
## year month day mean
## <int> <int> <int> <dbl>
## 1 2013 1 1 11.5
## 2 2013 1 2 13.9
## 3 2013 1 3 11.0
## 4 2013 1 4 8.95
## 5 2013 1 5 5.73
## 6 2013 1 6 7.15
## 7 2013 1 7 5.42
## 8 2013 1 8 2.55
## 9 2013 1 9 2.28
## 10 2013 1 10 2.84
## # ℹ 355 more rowsIn this case, where missings represent cancelled flights, we could first remove the cancelled flights.
not_cancelled <- flights %>%
filter(!is.na(dep_delay), !is.na(arr_delay))
not_cancelled %>%
group_by(year, month, day) %>%
summarise(mean = mean(dep_delay))## `summarise()` has grouped output by 'year', 'month'. You can override using the
## `.groups` argument.## # A tibble: 365 × 4
## # Groups: year, month [12]
## year month day mean
## <int> <int> <int> <dbl>
## 1 2013 1 1 11.4
## 2 2013 1 2 13.7
## 3 2013 1 3 10.9
## 4 2013 1 4 8.97
## 5 2013 1 5 5.73
## 6 2013 1 6 7.15
## 7 2013 1 7 5.42
## 8 2013 1 8 2.56
## 9 2013 1 9 2.30
## 10 2013 1 10 2.84
## # ℹ 355 more rowsCounts
Whenever you do any aggregation, it’s a good idea to include either a count (n()) or count of non-issing values (sum(!is.na(x))). This ensures you can check that you are not drawing conclusions based on very small amounts of data.
Ex: Look at the planes, identified by tail number, that have the highest average delays:
delays <- not_cancelled %>%
group_by(tailnum) %>%
summarise(
delay = mean(arr_delay)
)
ggplot(data = delays, mapping = aes(x = delay)) +
geom_freqpoly(binwidth = 10)
We can see that some planes have an average delay of 5 hours (300 minutes).
BUT, we can get more insight with a scatter plor of number of flights vs. average delay:
delays <- not_cancelled %>%
group_by(tailnum) %>%
summarise(
delay = mean(arr_delay, na.rm = TRUE),
n = n()
)
ggplot(data = delays, mapping = aes(x = n, y = delay)) +
geom_point(alpha = 1/10)
We can see that there is much greater variation in the average delay when there are only a few flights. This is a very common plot shape when mean (or some other summary) is plotted vs. group size. The variation decreases as the sample size increases.
With a plot like this, it’s useful to filter out the groups with the smallest number of observations so you can see more of a pattern and less extreme variation in the smaller groups.
# ggplot does not handle pipes, that is why '+' is used instead of '%>%'
delays %>%
filter(n > 25) %>%
ggplot(mapping = aes(x = n, y = delay)) +
geom_point(alpha = 1/10)
Another common variation of this type of pattern as can be seen in the Lahman baseball package: 1. As above, the variation of the aggregate decreases as we get more data points. 2. There’s a positive correlation between skill(ba) and opportunities to hit the ball(ab) This is because the teams control who plays and will pick their best players.
# Convert to a tibble so it prints nicely
library(Lahman)
batting <- as_tibble(Lahman::Batting)
batters <- batting %>%
group_by(playerID) %>%
summarise(
ba = sum(H, na.rm = TRUE) / sum(AB, na.rm = TRUE),
ab = sum(AB, na.rm = TRUE)
)
batters %>%
filter(ab > 100) %>%
ggplot(mapping = aes(x = ab, y = ba)) +
geom_point() +
geom_smooth(se = FALSE)## `geom_smooth()` using method = 'gam' and formula = 'y ~ s(x, bs = "cs")'
This has important implications for ranking. If you naively sort on desc(ba), the people with the best batting averages are clearly lucky, not skilled. (Only had 1 chance to hit and did get a hit).
batters %>%
arrange(desc(ba))## # A tibble: 20,469 × 3
## playerID ba ab
## <chr> <dbl> <int>
## 1 abramge01 1 1
## 2 alberan01 1 1
## 3 banisje01 1 1
## 4 bartocl01 1 1
## 5 bassdo01 1 1
## 6 birasst01 1 2
## 7 bruneju01 1 1
## 8 burnscb01 1 1
## 9 cammaer01 1 1
## 10 campsh01 1 1
## # ℹ 20,459 more rowsUseful Summary Functions
Besides means, counts, and sum…
- Measures of Location:
- mean(x): Mean = sum/length
- median(x): Median = A value where 50% of x is above and 50% below
not_cancelled %>%
group_by(year, month, day) %>%
summarise(
avg_delay1 = mean(arr_delay),
avg_delay2 = mean(arr_delay[arr_delay > 0]) # the average positive delay
)## `summarise()` has grouped output by 'year', 'month'. You can override using the
## `.groups` argument.## # A tibble: 365 × 5
## # Groups: year, month [12]
## year month day avg_delay1 avg_delay2
## <int> <int> <int> <dbl> <dbl>
## 1 2013 1 1 12.7 32.5
## 2 2013 1 2 12.7 32.0
## 3 2013 1 3 5.73 27.7
## 4 2013 1 4 -1.93 28.3
## 5 2013 1 5 -1.53 22.6
## 6 2013 1 6 4.24 24.4
## 7 2013 1 7 -4.95 27.8
## 8 2013 1 8 -3.23 20.8
## 9 2013 1 9 -0.264 25.6
## 10 2013 1 10 -5.90 27.3
## # ℹ 355 more rows- Measures of Spread:
- sd(x): Standard Deviation, or root mean squared deviation. The standard measure of spread.
- IQR(x): The Interquartile Range
- mad(x): Median Absolute Deviation
# Why is distance to some destinations more variable than to others?
not_cancelled %>%
group_by(dest) %>%
summarise(distance_sd = sd(distance)) %>%
arrange(desc(distance_sd))## # A tibble: 104 × 2
## dest distance_sd
## <chr> <dbl>
## 1 EGE 10.5
## 2 SAN 10.4
## 3 SFO 10.2
## 4 HNL 10.0
## 5 SEA 9.98
## 6 LAS 9.91
## 7 PDX 9.87
## 8 PHX 9.86
## 9 LAX 9.66
## 10 IND 9.46
## # ℹ 94 more rows- Measures of Rank:
- min(x)
- quantile(x, 0.25): Quantiles are the generalization of the median. quantile(x,0.25) finds the value of x that is greater than 25% of the values and less than the other 75%.
- max(x) Median Absolute Deviation
# When do the first and last flights leave each day?
not_cancelled %>%
group_by(year, month, day) %>%
summarise(
first = min(dep_time),
last = max(dep_time)
)## `summarise()` has grouped output by 'year', 'month'. You can override using the
## `.groups` argument.## # A tibble: 365 × 5
## # Groups: year, month [12]
## year month day first last
## <int> <int> <int> <int> <int>
## 1 2013 1 1 517 2356
## 2 2013 1 2 42 2354
## 3 2013 1 3 32 2349
## 4 2013 1 4 25 2358
## 5 2013 1 5 14 2357
## 6 2013 1 6 16 2355
## 7 2013 1 7 49 2359
## 8 2013 1 8 454 2351
## 9 2013 1 9 2 2252
## 10 2013 1 10 3 2320
## # ℹ 355 more rows- Measures of Position:
- first(x)
- nth(x, 2)
- last(x)
# Find the first and last departure of each day
not_cancelled %>%
group_by(year, month, day) %>%
summarise(
first_dep = first(dep_time),
last_dep = last(dep_time)
)## `summarise()` has grouped output by 'year', 'month'. You can override using the
## `.groups` argument.## # A tibble: 365 × 5
## # Groups: year, month [12]
## year month day first_dep last_dep
## <int> <int> <int> <int> <int>
## 1 2013 1 1 517 2356
## 2 2013 1 2 42 2354
## 3 2013 1 3 32 2349
## 4 2013 1 4 25 2358
## 5 2013 1 5 14 2357
## 6 2013 1 6 16 2355
## 7 2013 1 7 49 2359
## 8 2013 1 8 454 2351
## 9 2013 1 9 2 2252
## 10 2013 1 10 3 2320
## # ℹ 355 more rowsThese functions are complementary to filtering on ranks. Filtering gives you all variables, with each observation in a separate row:
not_cancelled %>%
group_by(year, month, day) %>%
mutate(r = min_rank(desc(dep_time))) %>%
filter(r %in% range(r))## # A tibble: 770 × 20
## # Groups: year, month, day [365]
## year month day dep_time sched_dep_time dep_delay arr_time sched_arr_time
## <int> <int> <int> <int> <int> <dbl> <int> <int>
## 1 2013 1 1 517 515 2 830 819
## 2 2013 1 1 2356 2359 -3 425 437
## 3 2013 1 2 42 2359 43 518 442
## 4 2013 1 2 2354 2359 -5 413 437
## 5 2013 1 3 32 2359 33 504 442
## 6 2013 1 3 2349 2359 -10 434 445
## 7 2013 1 4 25 2359 26 505 442
## 8 2013 1 4 2358 2359 -1 429 437
## 9 2013 1 4 2358 2359 -1 436 445
## 10 2013 1 5 14 2359 15 503 445
## # ℹ 760 more rows
## # ℹ 12 more variables: arr_delay <dbl>, carrier <chr>, flight <int>,
## # tailnum <chr>, origin <chr>, dest <chr>, air_time <dbl>, distance <dbl>,
## # hour <dbl>, minute <dbl>, time_hour <dttm>, r <int>- Counts:
- n(): Takes no arguments, returns the size of the current group.
- sum(!is.na(x)): Counts the number of non-missing values.
- n_distinct(x): Counts the number of distinct (unique) values.
# Which destinations have the most carriers?
not_cancelled %>%
group_by(dest) %>%
summarise(carriers = n_distinct(carrier)) %>%
arrange(desc(carriers))## # A tibble: 104 × 2
## dest carriers
## <chr> <int>
## 1 ATL 7
## 2 BOS 7
## 3 CLT 7
## 4 ORD 7
## 5 TPA 7
## 6 AUS 6
## 7 DCA 6
## 8 DTW 6
## 9 IAD 6
## 10 MSP 6
## # ℹ 94 more rows# dplyr provides a simple helper if all you want is count:
not_cancelled %>%
count(dest)## # A tibble: 104 × 2
## dest n
## <chr> <int>
## 1 ABQ 254
## 2 ACK 264
## 3 ALB 418
## 4 ANC 8
## 5 ATL 16837
## 6 AUS 2411
## 7 AVL 261
## 8 BDL 412
## 9 BGR 358
## 10 BHM 269
## # ℹ 94 more rows# You can optionally provide a weight variable. Ex: Count() or sum the total number of miles a plane flew:
not_cancelled %>%
count(tailnum, wt = distance)## # A tibble: 4,037 × 2
## tailnum n
## <chr> <dbl>
## 1 D942DN 3418
## 2 N0EGMQ 239143
## 3 N10156 109664
## 4 N102UW 25722
## 5 N103US 24619
## 6 N104UW 24616
## 7 N10575 139903
## 8 N105UW 23618
## 9 N107US 21677
## 10 N108UW 32070
## # ℹ 4,027 more rows- Counts and Proportions of Logical values:
- sum(x > 10): Gives the number of TRUEs in x
- mean(y = 0): Givest the Proportion of TRUEs in x
When used with numerics functions, TRUE is converted to 1 and FALSE to 0. This makes sum() ad mean() very useful.
# How many flights left before 5am? (these usually indicate delayed
# flights from the previous day)
not_cancelled %>%
group_by(year, month, day) %>%
summarise(n_early = sum(dep_time < 500))## `summarise()` has grouped output by 'year', 'month'. You can override using the
## `.groups` argument.## # A tibble: 365 × 4
## # Groups: year, month [12]
## year month day n_early
## <int> <int> <int> <int>
## 1 2013 1 1 0
## 2 2013 1 2 3
## 3 2013 1 3 4
## 4 2013 1 4 3
## 5 2013 1 5 3
## 6 2013 1 6 2
## 7 2013 1 7 2
## 8 2013 1 8 1
## 9 2013 1 9 3
## 10 2013 1 10 3
## # ℹ 355 more rows# What proportion of flights are delayed by more than an hour?
not_cancelled %>%
group_by(year, month, day) %>%
summarise(hour_prop = mean(arr_delay > 60))## `summarise()` has grouped output by 'year', 'month'. You can override using the
## `.groups` argument.## # A tibble: 365 × 4
## # Groups: year, month [12]
## year month day hour_prop
## <int> <int> <int> <dbl>
## 1 2013 1 1 0.0722
## 2 2013 1 2 0.0851
## 3 2013 1 3 0.0567
## 4 2013 1 4 0.0396
## 5 2013 1 5 0.0349
## 6 2013 1 6 0.0470
## 7 2013 1 7 0.0333
## 8 2013 1 8 0.0213
## 9 2013 1 9 0.0202
## 10 2013 1 10 0.0183
## # ℹ 355 more rowsGrouping by Multiple Variables
When you group_by() multiple variables, each summary peels off 1 level of grouping, making it easy to progressively roll up a dataset.
daily <- group_by(flights, year, month, day)
(per_day <- summarise(daily, flights = n()))## `summarise()` has grouped output by 'year', 'month'. You can override using the
## `.groups` argument.## # A tibble: 365 × 4
## # Groups: year, month [12]
## year month day flights
## <int> <int> <int> <int>
## 1 2013 1 1 842
## 2 2013 1 2 943
## 3 2013 1 3 914
## 4 2013 1 4 915
## 5 2013 1 5 720
## 6 2013 1 6 832
## 7 2013 1 7 933
## 8 2013 1 8 899
## 9 2013 1 9 902
## 10 2013 1 10 932
## # ℹ 355 more rows(per_month <- summarise(per_day, flights = sum(flights)))## `summarise()` has grouped output by 'year'. You can override using the
## `.groups` argument.## # A tibble: 12 × 3
## # Groups: year [1]
## year month flights
## <int> <int> <int>
## 1 2013 1 27004
## 2 2013 2 24951
## 3 2013 3 28834
## 4 2013 4 28330
## 5 2013 5 28796
## 6 2013 6 28243
## 7 2013 7 29425
## 8 2013 8 29327
## 9 2013 9 27574
## 10 2013 10 28889
## 11 2013 11 27268
## 12 2013 12 28135(per_year <- summarise(per_month, flights = sum(flights)))## # A tibble: 1 × 2
## year flights
## <int> <int>
## 1 2013 336776Note: Be careful when progressively rooling up summaries: it’s OK for counts and sums but stop and think about weighting means and variances. It’s not possible to do it exactly for rank-based stats like median. In other words, the sum of groupwise sums is the overall sum, but the median of groupwise medians is not the overall median.
Ungrouping
If you need to remove grouping, and return to operations on ungrouped data, use ungroup().
daily %>%
ungroup() %>% # no longer grouped by date
summarise(flights = n()) # all flights## # A tibble: 1 × 1
## flights
## <int>
## 1 3367765.7 Grouped Mutates and Filters
Grouping is most useful in conjunction with summarise(), but you can also do convenient operations with mutate() and filter(). Functions that work most naturally in grouped mutates and filters are known as window functions (vs the summary functions used for summaries).
- Grouped filter: A grouped mutate followed by an ungrouped filter. Generally best to avoid except for quick and dirty manipulations, otherwise hard to check it’s done correctly.
Find the worst members of each group:
flights_sml %>%
group_by(year, month, day) %>%
filter(rank(desc(arr_delay)) < 10)## # A tibble: 3,306 × 7
## # Groups: year, month, day [365]
## year month day dep_delay arr_delay distance air_time
## <int> <int> <int> <dbl> <dbl> <dbl> <dbl>
## 1 2013 1 1 853 851 184 41
## 2 2013 1 1 290 338 1134 213
## 3 2013 1 1 260 263 266 46
## 4 2013 1 1 157 174 213 60
## 5 2013 1 1 216 222 708 121
## 6 2013 1 1 255 250 589 115
## 7 2013 1 1 285 246 1085 146
## 8 2013 1 1 192 191 199 44
## 9 2013 1 1 379 456 1092 222
## 10 2013 1 2 224 207 550 94
## # ℹ 3,296 more rowsFind groups bigger than a threshold:
popular_dests <- flights %>%
group_by(dest) %>%
filter(n() > 365)
popular_dests## # A tibble: 332,577 × 19
## # Groups: dest [77]
## year month day dep_time sched_dep_time dep_delay arr_time sched_arr_time
## <int> <int> <int> <int> <int> <dbl> <int> <int>
## 1 2013 1 1 517 515 2 830 819
## 2 2013 1 1 533 529 4 850 830
## 3 2013 1 1 542 540 2 923 850
## 4 2013 1 1 544 545 -1 1004 1022
## 5 2013 1 1 554 600 -6 812 837
## 6 2013 1 1 554 558 -4 740 728
## 7 2013 1 1 555 600 -5 913 854
## 8 2013 1 1 557 600 -3 709 723
## 9 2013 1 1 557 600 -3 838 846
## 10 2013 1 1 558 600 -2 753 745
## # ℹ 332,567 more rows
## # ℹ 11 more variables: arr_delay <dbl>, carrier <chr>, flight <int>,
## # tailnum <chr>, origin <chr>, dest <chr>, air_time <dbl>, distance <dbl>,
## # hour <dbl>, minute <dbl>, time_hour <dttm>Standardise to compute per group metrics:
popular_dests %>%
filter(arr_delay > 0) %>%
mutate(prop_delay = arr_delay / sum(arr_delay)) %>%
select(year:day, dest, arr_delay, prop_delay)## # A tibble: 131,106 × 6
## # Groups: dest [77]
## year month day dest arr_delay prop_delay
## <int> <int> <int> <chr> <dbl> <dbl>
## 1 2013 1 1 IAH 11 0.000111
## 2 2013 1 1 IAH 20 0.000201
## 3 2013 1 1 MIA 33 0.000235
## 4 2013 1 1 ORD 12 0.0000424
## 5 2013 1 1 FLL 19 0.0000938
## 6 2013 1 1 ORD 8 0.0000283
## 7 2013 1 1 LAX 7 0.0000344
## 8 2013 1 1 DFW 31 0.000282
## 9 2013 1 1 ATL 12 0.0000400
## 10 2013 1 1 DTW 16 0.000116
## # ℹ 131,096 more rows