psyntur, Andrews and Roeser 2021); cognitive psychology; language acquisition
tidyverse (50%)ggplot2 (40%)Why using R (or code in general) to handle data?
** What do you think? **
Why using R (or code in general) to handle data?
.Rproj with one director per project with sub-directories for scripts, data, plots, etc# Ceci n'est pas un comment!)tidyverse instead of base R.Code > Download ZIP > unzip directory on your machine.hallam-r-workshop.Rprojwrangling/exercises/: exercises associated with each topicdata/: scripts read data from herewrangling/slides.Rmd: these slides in R markdown format (.html format provided as well)tidyverseCollection of R packages for data science that share:
tidyverse# Installs 19 packages
install.packages("tidyverse")
# Loads 6 packages library(tidyverse)
tidyverse
tidyverse: verbsFunctions that do specific things to our data.
Must know: read_csv, write_csv, glimpse, select, filter, mutate, group_by / ungroup, summarise, pivot_wider / _longer, _join, bind_rows / _cols
Also important: count, pull, slice, across, recode, unique, n, where, everything, ~ and ., map, starts_with, ends_with, contains, separate, unite, transmute
There are more but these are the most important ones.
blomkvist <- read_csv("../data/blomkvist.csv")
glimpse(blomkvist)
Rows: 354 Columns: 11 $ id <dbl> 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12,… $ sex <chr> "male", "female", "female", "female", … $ age <dbl> 84, 37, 62, 85, 73, 65, 30, 49, 83, 58… $ medicine <dbl> 8, 1, 0, 4, 5, 0, 0, 0, 11, 0, 0, 4, 3… $ smoker <chr> "former", "no", "yes", "former", "form… $ pal_work <dbl> NA, 2, NA, NA, NA, 1, 3, 1, NA, 4, 2, … $ pal_leisure <dbl> 1, 2, 2, 2, 3, 3, 2, 2, 1, 3, 3, 2, 1,… $ rt_hand_d <dbl> 702, 471, 639, 708, 607, 542, 571, 509… $ rt_hand_nd <dbl> 780, 497, 638, 639, 652, 499, 527, 547… $ rt_foot_d <dbl> 1009, 738, 878, 902, 923, 687, 778, 74… $ rt_foot_nd <dbl> 963, 692, 786, 1374, 805, 600, 750, 79…
rt) of dominant (_d) or non-dominant (_nd) hand or foot in msecsmedicine: number of drugs used dailypal: physical activity level: 1 (least) to 4 (most active)tidyverse is operating with tibbles# Imports data as data frame
data_as_frame <- read.csv("path_to_data/data.csv")
# Imports data as tibble
data_as_tibble <- read_csv("path_to_data/data.csv")
.csv: comma separated filereadr package: e.g. read_csv, read_delim, read_tsvFor other data formats:
haven package: e.g. read_dta, read_sav, read_sasreadxl package: e.g. read_excel, read_xls, read_xlsx# Summarise data structure in base R str(data_as_frame) # Summarise data structure in tidyverse glimpse(data_as_tibble)
Open exercise script 1
tidyverse functionsFunctions follow the principle
function_name(data_name, argument)
where argument specifies which variable / condition etc. the function has to operate on.
# Picking out variables select(data, variable1) # Subsetting data filter(data, variable > 100) # Change / add variables mutate(data, variable_sqr = variable^2) # Aggregate data summarise(data, mean_var = mean(variable))
Extracts variables you’re interested in.
glimpse(blomkvist)
Rows: 354 Columns: 11 $ id <dbl> 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17,… $ sex <chr> "male", "female", "female", "female", "male", "male", "fem… $ age <dbl> 84, 37, 62, 85, 73, 65, 30, 49, 83, 58, 25, 88, 62, 88, 27… $ medicine <dbl> 8, 1, 0, 4, 5, 0, 0, 0, 11, 0, 0, 4, 3, 8, 1, 3, 4, 1, 1, … $ smoker <chr> "former", "no", "yes", "former", "former", "no", "no", "fo… $ pal_work <dbl> NA, 2, NA, NA, NA, 1, 3, 1, NA, 4, 2, NA, 3, NA, 2, 3, NA,… $ pal_leisure <dbl> 1, 2, 2, 2, 3, 3, 2, 2, 1, 3, 3, 2, 1, 1, 3, 3, 1, 3, 1, 2… $ rt_hand_d <dbl> 702, 471, 639, 708, 607, 542, 571, 509, 737, 550, 548, 889… $ rt_hand_nd <dbl> 780, 497, 638, 639, 652, 499, 527, 547, 865, 569, 507, 766… $ rt_foot_d <dbl> 1009, 738, 878, 902, 923, 687, 778, 743, 750, 629, 653, 79… $ rt_foot_nd <dbl> 963, 692, 786, 1374, 805, 600, 750, 797, 797, 800, 718, 86…
Extracts variables you’re interested in.
glimpse(blomkvist)
Rows: 354 Columns: 11 $ id <dbl> 1, 2, 3,… $ sex <chr> "male", … $ age <dbl> 84, 37, … $ medicine <dbl> 8, 1, 0,… $ smoker <chr> "former"… $ pal_work <dbl> NA, 2, N… $ pal_leisure <dbl> 1, 2, 2,… $ rt_hand_d <dbl> 702, 471… $ rt_hand_nd <dbl> 780, 497… $ rt_foot_d <dbl> 1009, 73… $ rt_foot_nd <dbl> 963, 692…
select(blomkvist, id, sex, age)
# A tibble: 354 × 3
id sex age
<dbl> <chr> <dbl>
1 1 male 84
2 2 female 37
3 3 female 62
4 4 female 85
5 5 male 73
# … with 349 more rowsExtracts variables you’re interested in.
glimpse(blomkvist)
Rows: 354 Columns: 11 $ id <dbl> 1, 2, 3,… $ sex <chr> "male", … $ age <dbl> 84, 37, … $ medicine <dbl> 8, 1, 0,… $ smoker <chr> "former"… $ pal_work <dbl> NA, 2, N… $ pal_leisure <dbl> 1, 2, 2,… $ rt_hand_d <dbl> 702, 471… $ rt_hand_nd <dbl> 780, 497… $ rt_foot_d <dbl> 1009, 73… $ rt_foot_nd <dbl> 963, 692…
select(blomkvist, id:age)
# A tibble: 354 × 3
id sex age
<dbl> <chr> <dbl>
1 1 male 84
2 2 female 37
3 3 female 62
4 4 female 85
5 5 male 73
# … with 349 more rowsExtracts variables you’re interested in.
glimpse(blomkvist)
Rows: 354 Columns: 11 $ id <dbl> 1, 2, 3,… $ sex <chr> "male", … $ age <dbl> 84, 37, … $ medicine <dbl> 8, 1, 0,… $ smoker <chr> "former"… $ pal_work <dbl> NA, 2, N… $ pal_leisure <dbl> 1, 2, 2,… $ rt_hand_d <dbl> 702, 471… $ rt_hand_nd <dbl> 780, 497… $ rt_foot_d <dbl> 1009, 73… $ rt_foot_nd <dbl> 963, 692…
select(blomkvist, 1, 2, 3)
# A tibble: 354 × 3
id sex age
<dbl> <chr> <dbl>
1 1 male 84
2 2 female 37
3 3 female 62
4 4 female 85
5 5 male 73
# … with 349 more rowsExtracts variables you’re interested in.
glimpse(blomkvist)
Rows: 354 Columns: 11 $ id <dbl> 1, 2, 3,… $ sex <chr> "male", … $ age <dbl> 84, 37, … $ medicine <dbl> 8, 1, 0,… $ smoker <chr> "former"… $ pal_work <dbl> NA, 2, N… $ pal_leisure <dbl> 1, 2, 2,… $ rt_hand_d <dbl> 702, 471… $ rt_hand_nd <dbl> 780, 497… $ rt_foot_d <dbl> 1009, 73… $ rt_foot_nd <dbl> 963, 692…
select(blomkvist, 1:3)
# A tibble: 354 × 3
id sex age
<dbl> <chr> <dbl>
1 1 male 84
2 2 female 37
3 3 female 62
4 4 female 85
5 5 male 73
# … with 349 more rowsExtracts variables you’re interested in.
glimpse(blomkvist)
Rows: 354 Columns: 11 $ id <dbl> 1, 2, 3,… $ sex <chr> "male", … $ age <dbl> 84, 37, … $ medicine <dbl> 8, 1, 0,… $ smoker <chr> "former"… $ pal_work <dbl> NA, 2, N… $ pal_leisure <dbl> 1, 2, 2,… $ rt_hand_d <dbl> 702, 471… $ rt_hand_nd <dbl> 780, 497… $ rt_foot_d <dbl> 1009, 73… $ rt_foot_nd <dbl> 963, 692…
select(blomkvist, id, sex, rt = rt_hand_d)
# A tibble: 354 × 3
id sex rt
<dbl> <chr> <dbl>
1 1 male 702.
2 2 female 471.
3 3 female 639.
4 4 female 708
5 5 male 607.
# … with 349 more rowsExtracts variables you’re interested in.
glimpse(blomkvist)
Rows: 354 Columns: 11 $ id <dbl> 1, 2, 3,… $ sex <chr> "male", … $ age <dbl> 84, 37, … $ medicine <dbl> 8, 1, 0,… $ smoker <chr> "former"… $ pal_work <dbl> NA, 2, N… $ pal_leisure <dbl> 1, 2, 2,… $ rt_hand_d <dbl> 702, 471… $ rt_hand_nd <dbl> 780, 497… $ rt_foot_d <dbl> 1009, 73… $ rt_foot_nd <dbl> 963, 692…
select(blomkvist, id, starts_with("rt_"))
# A tibble: 354 × 5
id rt_hand_d rt_hand_nd rt_foot_d rt_foot_nd
<dbl> <dbl> <dbl> <dbl> <dbl>
1 1 702. 780. 1009 963.
2 2 471. 497 738. 692.
3 3 639. 638 878 786
4 4 708 639. 902. 1374.
5 5 607. 652 923 805
# … with 349 more rowsExtracts variables you’re interested in.
glimpse(blomkvist)
Rows: 354 Columns: 11 $ id <dbl> 1, 2, 3,… $ sex <chr> "male", … $ age <dbl> 84, 37, … $ medicine <dbl> 8, 1, 0,… $ smoker <chr> "former"… $ pal_work <dbl> NA, 2, N… $ pal_leisure <dbl> 1, 2, 2,… $ rt_hand_d <dbl> 702, 471… $ rt_hand_nd <dbl> 780, 497… $ rt_foot_d <dbl> 1009, 73… $ rt_foot_nd <dbl> 963, 692…
select(blomkvist, id, ends_with("d"))
???
Extracts variables you’re interested in.
glimpse(blomkvist)
Rows: 354 Columns: 11 $ id <dbl> 1, 2, 3,… $ sex <chr> "male", … $ age <dbl> 84, 37, … $ medicine <dbl> 8, 1, 0,… $ smoker <chr> "former"… $ pal_work <dbl> NA, 2, N… $ pal_leisure <dbl> 1, 2, 2,… $ rt_hand_d <dbl> 702, 471… $ rt_hand_nd <dbl> 780, 497… $ rt_foot_d <dbl> 1009, 73… $ rt_foot_nd <dbl> 963, 692…
select(blomkvist, id, ends_with("d"))
# A tibble: 354 × 5
id rt_hand_d rt_hand_nd rt_foot_d rt_foot_nd
<dbl> <dbl> <dbl> <dbl> <dbl>
1 1 702. 780. 1009 963.
2 2 471. 497 738. 692.
3 3 639. 638 878 786
4 4 708 639. 902. 1374.
5 5 607. 652 923 805
# … with 349 more rowsExtracts variables you’re interested in.
glimpse(blomkvist)
Rows: 354 Columns: 11 $ id <dbl> 1, 2, 3,… $ sex <chr> "male", … $ age <dbl> 84, 37, … $ medicine <dbl> 8, 1, 0,… $ smoker <chr> "former"… $ pal_work <dbl> NA, 2, N… $ pal_leisure <dbl> 1, 2, 2,… $ rt_hand_d <dbl> 702, 471… $ rt_hand_nd <dbl> 780, 497… $ rt_foot_d <dbl> 1009, 73… $ rt_foot_nd <dbl> 963, 692…
select(blomkvist, id, contains("hand"))
???
Extracts variables you’re interested in.
glimpse(blomkvist)
Rows: 354 Columns: 11 $ id <dbl> 1, 2, 3,… $ sex <chr> "male", … $ age <dbl> 84, 37, … $ medicine <dbl> 8, 1, 0,… $ smoker <chr> "former"… $ pal_work <dbl> NA, 2, N… $ pal_leisure <dbl> 1, 2, 2,… $ rt_hand_d <dbl> 702, 471… $ rt_hand_nd <dbl> 780, 497… $ rt_foot_d <dbl> 1009, 73… $ rt_foot_nd <dbl> 963, 692…
select(blomkvist, id, contains("hand"))
# A tibble: 354 × 3
id rt_hand_d rt_hand_nd
<dbl> <dbl> <dbl>
1 1 702. 780.
2 2 471. 497
3 3 639. 638
4 4 708 639.
5 5 607. 652
# … with 349 more rowsExtracts variables you’re interested in.
glimpse(blomkvist)
Rows: 354 Columns: 11 $ id <dbl> 1, 2, 3,… $ sex <chr> "male", … $ age <dbl> 84, 37, … $ medicine <dbl> 8, 1, 0,… $ smoker <chr> "former"… $ pal_work <dbl> NA, 2, N… $ pal_leisure <dbl> 1, 2, 2,… $ rt_hand_d <dbl> 702, 471… $ rt_hand_nd <dbl> 780, 497… $ rt_foot_d <dbl> 1009, 73… $ rt_foot_nd <dbl> 963, 692…
select(blomkvist, -contains("hand"))
# A tibble: 354 × 9
id sex age medicine smoker pal_work pal_leisure rt_foot_d rt_foot_nd
<dbl> <chr> <dbl> <dbl> <chr> <dbl> <dbl> <dbl> <dbl>
1 1 male 84 8 former NA 1 1009 963.
2 2 female 37 1 no 2 2 738. 692.
3 3 female 62 0 yes NA 2 878 786
4 4 female 85 4 former NA 2 902. 1374.
5 5 male 73 5 former NA 3 923 805
# … with 349 more rowsExtracts variables you’re interested in.
glimpse(blomkvist)
Rows: 354 Columns: 11 $ id <dbl> 1, 2, 3,… $ sex <chr> "male", … $ age <dbl> 84, 37, … $ medicine <dbl> 8, 1, 0,… $ smoker <chr> "former"… $ pal_work <dbl> NA, 2, N… $ pal_leisure <dbl> 1, 2, 2,… $ rt_hand_d <dbl> 702, 471… $ rt_hand_nd <dbl> 780, 497… $ rt_foot_d <dbl> 1009, 73… $ rt_foot_nd <dbl> 963, 692…
select(blomkvist, -ends_with("_d"))
???
Extracts variables you’re interested in.
glimpse(blomkvist)
Rows: 354 Columns: 11 $ id <dbl> 1, 2, 3,… $ sex <chr> "male", … $ age <dbl> 84, 37, … $ medicine <dbl> 8, 1, 0,… $ smoker <chr> "former"… $ pal_work <dbl> NA, 2, N… $ pal_leisure <dbl> 1, 2, 2,… $ rt_hand_d <dbl> 702, 471… $ rt_hand_nd <dbl> 780, 497… $ rt_foot_d <dbl> 1009, 73… $ rt_foot_nd <dbl> 963, 692…
select(blomkvist, -ends_with("_d"))
# A tibble: 354 × 9
id sex age medicine smoker pal_work pal_leisure rt_hand_nd rt_foot_nd
<dbl> <chr> <dbl> <dbl> <chr> <dbl> <dbl> <dbl> <dbl>
1 1 male 84 8 former NA 1 780. 963.
2 2 female 37 1 no 2 2 497 692.
3 3 female 62 0 yes NA 2 638 786
4 4 female 85 4 former NA 2 639. 1374.
5 5 male 73 5 former NA 3 652 805
# … with 349 more rowsExtracts variables you’re interested in.
glimpse(blomkvist)
Rows: 354 Columns: 11 $ id <dbl> 1, 2, 3,… $ sex <chr> "male", … $ age <dbl> 84, 37, … $ medicine <dbl> 8, 1, 0,… $ smoker <chr> "former"… $ pal_work <dbl> NA, 2, N… $ pal_leisure <dbl> 1, 2, 2,… $ rt_hand_d <dbl> 702, 471… $ rt_hand_nd <dbl> 780, 497… $ rt_foot_d <dbl> 1009, 73… $ rt_foot_nd <dbl> 963, 692…
select(blomkvist, -sex:-smoker)
???
Extracts variables you’re interested in.
glimpse(blomkvist)
Rows: 354 Columns: 11 $ id <dbl> 1, 2, 3,… $ sex <chr> "male", … $ age <dbl> 84, 37, … $ medicine <dbl> 8, 1, 0,… $ smoker <chr> "former"… $ pal_work <dbl> NA, 2, N… $ pal_leisure <dbl> 1, 2, 2,… $ rt_hand_d <dbl> 702, 471… $ rt_hand_nd <dbl> 780, 497… $ rt_foot_d <dbl> 1009, 73… $ rt_foot_nd <dbl> 963, 692…
select(blomkvist, -sex:-smoker)
# A tibble: 354 × 7
id pal_work pal_leisure rt_hand_d rt_hand_nd rt_foot_d rt_foot_nd
<dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl>
1 1 NA 1 702. 780. 1009 963.
2 2 2 2 471. 497 738. 692.
3 3 NA 2 639. 638 878 786
4 4 NA 2 708 639. 902. 1374.
5 5 NA 3 607. 652 923 805
# … with 349 more rowsExtracts variables you’re interested in.
glimpse(blomkvist)
Rows: 354 Columns: 11 $ id <dbl> 1, 2, 3,… $ sex <chr> "male", … $ age <dbl> 84, 37, … $ medicine <dbl> 8, 1, 0,… $ smoker <chr> "former"… $ pal_work <dbl> NA, 2, N… $ pal_leisure <dbl> 1, 2, 2,… $ rt_hand_d <dbl> 702, 471… $ rt_hand_nd <dbl> 780, 497… $ rt_foot_d <dbl> 1009, 73… $ rt_foot_nd <dbl> 963, 692…
select(blomkvist, where(is.character))
# A tibble: 354 × 2 sex smoker <chr> <chr> 1 male former 2 female no 3 female yes 4 female former 5 male former # … with 349 more rows
Extracts variables you’re interested in.
glimpse(blomkvist)
Rows: 354 Columns: 11 $ id <dbl> 1, 2, 3,… $ sex <chr> "male", … $ age <dbl> 84, 37, … $ medicine <dbl> 8, 1, 0,… $ smoker <chr> "former"… $ pal_work <dbl> NA, 2, N… $ pal_leisure <dbl> 1, 2, 2,… $ rt_hand_d <dbl> 702, 471… $ rt_hand_nd <dbl> 780, 497… $ rt_foot_d <dbl> 1009, 73… $ rt_foot_nd <dbl> 963, 692…
select(blomkvist, where(is.numeric))
???
Extracts variables you’re interested in.
glimpse(blomkvist)
Rows: 354 Columns: 11 $ id <dbl> 1, 2, 3,… $ sex <chr> "male", … $ age <dbl> 84, 37, … $ medicine <dbl> 8, 1, 0,… $ smoker <chr> "former"… $ pal_work <dbl> NA, 2, N… $ pal_leisure <dbl> 1, 2, 2,… $ rt_hand_d <dbl> 702, 471… $ rt_hand_nd <dbl> 780, 497… $ rt_foot_d <dbl> 1009, 73… $ rt_foot_nd <dbl> 963, 692…
select(blomkvist, where(is.numeric))
# A tibble: 354 × 9
id age medicine pal_work pal_leisure rt_hand_d rt_hand_nd rt_foot_d
<dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl>
1 1 84 8 NA 1 702. 780. 1009
2 2 37 1 2 2 471. 497 738.
3 3 62 0 NA 2 639. 638 878
4 4 85 4 NA 2 708 639. 902.
5 5 73 5 NA 3 607. 652 923
# … with 349 more rows, and 1 more variable: rt_foot_nd <dbl>
Continue with exercise 2
Select variables of interest
blomkvist_rt <- select(blomkvist, id, smoker, age, rt = rt_hand_d)
Subsetting data by selecting rows that meet one condition or more.
filter(data, condition)
filter(blomkvist_rt, rt >= 708)
# A tibble: 81 × 4
id smoker age rt
<dbl> <chr> <dbl> <dbl>
1 4 former 85 708
2 9 former 83 737.
3 12 no 88 889
4 13 yes 62 884.
5 14 former 88 832.
# … with 76 more rowsfilter(blomkvist_rt, rt > 708)
# A tibble: 79 × 4
id smoker age rt
<dbl> <chr> <dbl> <dbl>
1 9 former 83 737.
2 12 no 88 889
3 13 yes 62 884.
4 14 former 88 832.
5 17 former 80 930
# … with 74 more rowsfilter(blomkvist_rt, rt > 708, rt < 900)
# A tibble: 49 × 4
id smoker age rt
<dbl> <chr> <dbl> <dbl>
1 9 former 83 737.
2 12 no 88 889
3 13 yes 62 884.
4 14 former 88 832.
5 39 no 88 727
# … with 44 more rows
filter(blomkvist_rt, rt > 708 | rt < 900)
# A tibble: 353 × 4
id smoker age rt
<dbl> <chr> <dbl> <dbl>
1 1 former 84 702.
2 2 no 37 471.
3 3 yes 62 639.
4 4 former 85 708
5 5 former 73 607.
# … with 348 more rows
# which is a logical tautology really
filter(blomkvist_rt, rt < 708 | rt > 900)
# A tibble: 302 × 4
id smoker age rt
<dbl> <chr> <dbl> <dbl>
1 1 former 84 702.
2 2 no 37 471.
3 3 yes 62 639.
4 5 former 73 607.
5 6 no 65 542.
# … with 297 more rowsfilter(blomkvist_rt, rt > mean(rt, na.rm = TRUE))
# A tibble: 126 × 4
id smoker age rt
<dbl> <chr> <dbl> <dbl>
1 1 former 84 702.
2 3 yes 62 639.
3 4 former 85 708
4 9 former 83 737.
5 12 no 88 889
# … with 121 more rowsmean(blomkvist_rt$rt, na.rm = TRUE)
[1] 637
na.rm = TRUE?# Data with missing values y <- c(100, 1150, 200, 43, NA, 15)
mean(y)
[1] NA
mean(y, na.rm = TRUE)
[1] 302
sd(y, na.rm = TRUE)
[1] 480
filter(blomkvist_rt, smoker == "yes")
# A tibble: 30 × 4
id smoker age rt
<dbl> <chr> <dbl> <dbl>
1 3 yes 62 639.
2 10 yes 58 550.
3 13 yes 62 884.
4 24 yes 57 612.
5 28 yes 59 586.
# … with 25 more rowsunique(blomkvist_rt$smoker)
[1] "former" "no" "yes" NA
filter(blomkvist_rt, smoker != "yes")
# A tibble: 315 × 4
id smoker age rt
<dbl> <chr> <dbl> <dbl>
1 1 former 84 702.
2 2 no 37 471.
3 4 former 85 708
4 5 former 73 607.
5 6 no 65 542.
# … with 310 more rowsunique(blomkvist_rt$smoker)
[1] "former" "no" "yes" NA
filter(blomkvist_rt, smoker %in% c("yes", "former"))
# A tibble: 132 × 4
id smoker age rt
<dbl> <chr> <dbl> <dbl>
1 1 former 84 702.
2 3 yes 62 639.
3 4 former 85 708
4 5 former 73 607.
5 8 former 49 509.
# … with 127 more rowsfilter(blomkvist_rt, !(smoker %in% c("yes", "former")))
# A tibble: 222 × 4
id smoker age rt
<dbl> <chr> <dbl> <dbl>
1 2 no 37 471.
2 6 no 65 542.
3 7 no 30 571.
4 11 no 25 548
5 12 no 88 889
# … with 217 more rowsunique(blomkvist_rt$smoker)
[1] "former" "no" "yes" NA
filter(blomkvist_rt, smoker == "NA") # ooops!!!
# A tibble: 0 × 4 # … with 4 variables: id <dbl>, smoker <chr>, age <dbl>, rt <dbl>
filter(blomkvist_rt, smoker == NA) # double ooops!!!
# A tibble: 0 × 4 # … with 4 variables: id <dbl>, smoker <chr>, age <dbl>, rt <dbl>
filter(blomkvist_rt, is.na(smoker))
# A tibble: 9 × 4
id smoker age rt
<dbl> <chr> <dbl> <dbl>
1 41 <NA> 52 592
2 144 <NA> 36 501.
3 149 <NA> 34 479.
4 178 <NA> 26 466
5 183 <NA> 25 888.
# … with 4 more rowsfilter(blomkvist_rt, !is.na(smoker))
# A tibble: 345 × 4
id smoker age rt
<dbl> <chr> <dbl> <dbl>
1 1 former 84 702.
2 2 no 37 471.
3 3 yes 62 639.
4 4 former 85 708
5 5 former 73 607.
# … with 340 more rows
Continue with exercise 3
Adding new variables to the data or changing existing ones.
mutate(blomkvist_rt, rt_2 = rt * rt)
# A tibble: 354 × 5
id smoker age rt rt_2
<dbl> <chr> <dbl> <dbl> <dbl>
1 1 former 84 702. 492336.
2 2 no 37 471. 221527.
3 3 yes 62 639. 407895.
4 4 former 85 708 501264
5 5 former 73 607. 368854.
# … with 349 more rowsmutate(blomkvist_rt, rt_2 = rt^2)
# A tibble: 354 × 5
id smoker age rt rt_2
<dbl> <chr> <dbl> <dbl> <dbl>
1 1 former 84 702. 492336.
2 2 no 37 471. 221527.
3 3 yes 62 639. 407895.
4 4 former 85 708 501264
5 5 former 73 607. 368854.
# … with 349 more rowsmutate(blomkvist_rt, log_rt = log(rt))
# A tibble: 354 × 5
id smoker age rt log_rt
<dbl> <chr> <dbl> <dbl> <dbl>
1 1 former 84 702. 6.55
2 2 no 37 471. 6.15
3 3 yes 62 639. 6.46
4 4 former 85 708 6.56
5 5 former 73 607. 6.41
# … with 349 more rows
mutate(blomkvist_rt, is_slow = rt > 700)
# A tibble: 354 × 5
id smoker age rt is_slow
<dbl> <chr> <dbl> <dbl> <lgl>
1 1 former 84 702. TRUE
2 2 no 37 471. FALSE
3 3 yes 62 639. FALSE
4 4 former 85 708 TRUE
5 5 former 73 607. FALSE
# … with 349 more rows
mutate(blomkvist_rt, mean_rt = mean(rt, na.rm = TRUE))
# A tibble: 354 × 5
id smoker age rt mean_rt
<dbl> <chr> <dbl> <dbl> <dbl>
1 1 former 84 702. 637.
2 2 no 37 471. 637.
3 3 yes 62 639. 637.
4 4 former 85 708 637.
5 5 former 73 607. 637.
# … with 349 more rows
mutate(blomkvist_rt, mean_rt = mean(rt, na.rm = TRUE),
is_slow = rt > mean_rt)
# A tibble: 354 × 6
id smoker age rt mean_rt is_slow
<dbl> <chr> <dbl> <dbl> <dbl> <lgl>
1 1 former 84 702. 637. TRUE
2 2 no 37 471. 637. FALSE
3 3 yes 62 639. 637. TRUE
4 4 former 85 708 637. TRUE
5 5 former 73 607. 637. FALSE
# … with 349 more rowsmutate(blomkvist_rt, # or both in one go
is_slow = rt > mean(rt, na.rm = TRUE))recode()mutate(blomkvist_rt, smoker_recoded = recode(smoker, former = "former smoker",
yes = "smoker",
no = "non-smoker"))
# A tibble: 354 × 5
id smoker age rt smoker_recoded
<dbl> <chr> <dbl> <dbl> <chr>
1 1 former 84 702. former smoker
2 2 no 37 471. non-smoker
3 3 yes 62 639. smoker
4 4 former 85 708 former smoker
5 5 former 73 607. former smoker
# … with 349 more rows
cut()mutate(blomkvist_rt, age_cat = cut(age,
breaks = 3,
labels = c("low", "middle", "high")))
# A tibble: 354 × 5
id smoker age rt age_cat
<dbl> <chr> <dbl> <dbl> <fct>
1 1 former 84 702. high
2 2 no 37 471. low
3 3 yes 62 639. middle
4 4 former 85 708 high
5 5 former 73 607. high
# … with 349 more rows
case_when()case_when(condition ~ do) # similar to ifelse()
mutate(blomkvist_rt, age_cat = case_when(age > 70 ~ "high",
age > 40 ~ "middle",
is.na(smoker) ~ "dunno",
TRUE ~ "on the young side"))
# A tibble: 354 × 5
id smoker age rt age_cat
<dbl> <chr> <dbl> <dbl> <chr>
1 1 former 84 702. high
2 2 no 37 471. on the young side
3 3 yes 62 639. middle
4 4 former 85 708 high
5 5 former 73 607. high
# … with 349 more rows
Continue with exercise 4
group_by()Perform an action (function) for each level the grouping variable individually.
blomkvist_grouped <- group_by(blomkvist_rt, smoker) # Group by smoker blomkvist_grouped
# A tibble: 354 × 4
# Groups: smoker [4]
id smoker age rt
<dbl> <chr> <dbl> <dbl>
1 1 former 84 702.
2 2 no 37 471.
3 3 yes 62 639.
4 4 former 85 708
5 5 former 73 607.
# … with 349 more rows
# How many groups and what are they?
mutate(blomkvist_rt, # not grouped
mean_rt = mean(rt, na.rm = TRUE))
# A tibble: 354 × 5
id smoker age rt mean_rt
<dbl> <chr> <dbl> <dbl> <dbl>
1 1 former 84 702. 637.
2 2 no 37 471. 637.
3 3 yes 62 639. 637.
4 4 former 85 708 637.
5 5 former 73 607. 637.
# … with 349 more rowsmutate(blomkvist_grouped, # grouped data
mean_rt = mean(rt, na.rm = TRUE))
# A tibble: 354 × 5
# Groups: smoker [4]
id smoker age rt mean_rt
<dbl> <chr> <dbl> <dbl> <dbl>
1 1 former 84 702. 676.
2 2 no 37 471. 623.
3 3 yes 62 639. 628.
4 4 former 85 708 676.
5 5 former 73 607. 676.
# … with 349 more rows
# What's the grouping variable? # What's the difference?
otherwise you will keep performing by-group operations, even if you didn’t intend to.
blomkvist_grouped # :(
# A tibble: 354 × 4
# Groups: smoker [4]
id smoker age rt
<dbl> <chr> <dbl> <dbl>
1 1 former 84 702.
2 2 no 37 471.
3 3 yes 62 639.
4 4 former 85 708
5 5 former 73 607.
# … with 349 more rowsungroup(blomkvist_grouped) # :)
# A tibble: 354 × 4
id smoker age rt
<dbl> <chr> <dbl> <dbl>
1 1 former 84 702.
2 2 no 37 471.
3 3 yes 62 639.
4 4 former 85 708
5 5 former 73 607.
# … with 349 more rows
# Spot the difference!
… using descriptive tools.
summarise(blomkvist_rt, # not grouped
mean_rt = mean(rt, na.rm = TRUE),
N = n())
# A tibble: 1 × 2
mean_rt N
<dbl> <int>
1 637. 354summarise(blomkvist_grouped, # grouped data
mean_rt = mean(rt, na.rm = TRUE),
N = n())
# A tibble: 4 × 3 smoker mean_rt N <chr> <dbl> <int> 1 former 676. 102 2 no 623. 213 3 yes 628. 30 4 <NA> 549. 9
Summarising data using descriptive tools.
summarise(blomkvist_rt, # not grouped
mean_rt = mean(rt, na.rm = TRUE),
sd_rt = sd(rt, na.rm = TRUE),
min_rt = min(rt, na.rm = TRUE),
max_rt = max(rt, na.rm = TRUE),
N = n())
# A tibble: 1 × 5
mean_rt sd_rt min_rt max_rt N
<dbl> <dbl> <dbl> <dbl> <int>
1 637. 204. 379. 2076 354summarise(blomkvist_grouped, # grouped data
mean_rt = mean(rt, na.rm = TRUE),
sd_rt = sd(rt, na.rm = TRUE),
min_rt = min(rt, na.rm = TRUE),
max_rt = max(rt, na.rm = TRUE),
N = n())
# A tibble: 4 × 6 smoker mean_rt sd_rt min_rt max_rt N <chr> <dbl> <dbl> <dbl> <dbl> <int> 1 former 676. 231. 390. 2076 102 2 no 623. 192. 379. 1552. 213 3 yes 628. 202. 427. 1527 30 4 <NA> 549. 133. 466 888. 9
across()Create a new column.
mutate(blomkvist_rt, log_rt = log(rt))
# A tibble: 354 × 5
id smoker age rt log_rt
<dbl> <chr> <dbl> <dbl> <dbl>
1 1 former 84 702. 6.55
2 2 no 37 471. 6.15
3 3 yes 62 639. 6.46
4 4 former 85 708 6.56
5 5 former 73 607. 6.41
# … with 349 more rowsReplace with transformed variable.
mutate(blomkvist_rt, across(rt, log))
# A tibble: 354 × 4
id smoker age rt
<dbl> <chr> <dbl> <dbl>
1 1 former 84 6.55
2 2 no 37 6.15
3 3 yes 62 6.46
4 4 former 85 6.56
5 5 former 73 6.41
# … with 349 more rows
First argument of across will be used as first argument of function (log) which is supplied as second argument to across.
across()# Instead of
mutate(blomkvist, rt_hand_d = log(rt_hand_d),
rt_hand_nd = log(rt_hand_nd))
# Do mutate(blomkvist, across(c(rt_hand_d, rt_hand_nd), log))
# A tibble: 342 × 7
id age smoker rt_hand_d rt_hand_nd rt_foot_d rt_foot_nd
<dbl> <dbl> <chr> <dbl> <dbl> <dbl> <dbl>
1 1 84 former 6.55 6.66 1009 963.
2 2 37 no 6.15 6.21 738. 692.
3 3 62 yes 6.46 6.46 878 786
4 4 85 former 6.56 6.46 902. 1374.
5 5 73 former 6.41 6.48 923 805
# … with 337 more rows
across()mutate(blomkvist, across(where(is.numeric), log))
# A tibble: 342 × 7
id age smoker rt_hand_d rt_hand_nd rt_foot_d rt_foot_nd
<dbl> <dbl> <chr> <dbl> <dbl> <dbl> <dbl>
1 0 4.43 former 6.55 6.66 6.92 6.87
2 0.693 3.61 no 6.15 6.21 6.60 6.54
3 1.10 4.13 yes 6.46 6.46 6.78 6.67
4 1.39 4.44 former 6.56 6.46 6.80 7.23
5 1.61 4.29 former 6.41 6.48 6.83 6.69
# … with 337 more rows
# Not ideal here: see `id` column
across()mutate(blomkvist, across(starts_with("rt_"), log))
Rows: 342 Columns: 7 $ id <dbl> 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 15, 16, 17, 18, … $ age <dbl> 84, 37, 62, 85, 73, 65, 30, 49, 83, 58, 25, 88, 62, 27, 60,… $ smoker <chr> "former", "no", "yes", "former", "former", "no", "no", "for… $ rt_hand_d <dbl> 6.6, 6.2, 6.5, 6.6, 6.4, 6.3, 6.3, 6.2, 6.6, 6.3, 6.3, 6.8,… $ rt_hand_nd <dbl> 6.7, 6.2, 6.5, 6.5, 6.5, 6.2, 6.3, 6.3, 6.8, 6.3, 6.2, 6.6,… $ rt_foot_d <dbl> 6.9, 6.6, 6.8, 6.8, 6.8, 6.5, 6.7, 6.6, 6.6, 6.4, 6.5, 6.7,… $ rt_foot_nd <dbl> 6.9, 6.5, 6.7, 7.2, 6.7, 6.4, 6.6, 6.7, 6.7, 6.7, 6.6, 6.8,…
across()mutate(blomkvist, across(starts_with("rt_"), log, .names = "log_{.col}"))
Rows: 342 Columns: 11 $ id <dbl> 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 15, 16, 17, … $ age <dbl> 84, 37, 62, 85, 73, 65, 30, 49, 83, 58, 25, 88, 62, 27,… $ smoker <chr> "former", "no", "yes", "former", "former", "no", "no", … $ rt_hand_d <dbl> 702, 471, 639, 708, 607, 542, 571, 509, 737, 550, 548, … $ rt_hand_nd <dbl> 780, 497, 638, 639, 652, 499, 527, 547, 865, 569, 507, … $ rt_foot_d <dbl> 1009, 738, 878, 902, 923, 687, 778, 743, 750, 629, 653,… $ rt_foot_nd <dbl> 963, 692, 786, 1374, 805, 600, 750, 797, 797, 800, 718,… $ log_rt_hand_d <dbl> 6.6, 6.2, 6.5, 6.6, 6.4, 6.3, 6.3, 6.2, 6.6, 6.3, 6.3, … $ log_rt_hand_nd <dbl> 6.7, 6.2, 6.5, 6.5, 6.5, 6.2, 6.3, 6.3, 6.8, 6.3, 6.2, … $ log_rt_foot_d <dbl> 6.9, 6.6, 6.8, 6.8, 6.8, 6.5, 6.7, 6.6, 6.6, 6.4, 6.5, … $ log_rt_foot_nd <dbl> 6.9, 6.5, 6.7, 7.2, 6.7, 6.4, 6.6, 6.7, 6.7, 6.7, 6.6, …
across()mutate(blomkvist, across(starts_with("rt_"), list(lg = log, sqr = ~.^2), .names = "{.fn}_{.col}"))
Rows: 342 Columns: 15 $ id <dbl> 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 15, 16, 17, … $ age <dbl> 84, 37, 62, 85, 73, 65, 30, 49, 83, 58, 25, 88, 62, 27,… $ smoker <chr> "former", "no", "yes", "former", "former", "no", "no", … $ rt_hand_d <dbl> 702, 471, 639, 708, 607, 542, 571, 509, 737, 550, 548, … $ rt_hand_nd <dbl> 780, 497, 638, 639, 652, 499, 527, 547, 865, 569, 507, … $ rt_foot_d <dbl> 1009, 738, 878, 902, 923, 687, 778, 743, 750, 629, 653,… $ rt_foot_nd <dbl> 963, 692, 786, 1374, 805, 600, 750, 797, 797, 800, 718,… $ lg_rt_hand_d <dbl> 6.6, 6.2, 6.5, 6.6, 6.4, 6.3, 6.3, 6.2, 6.6, 6.3, 6.3, … $ sqr_rt_hand_d <dbl> 492336, 221527, 407895, 501264, 368854, 293403, 325660,… $ lg_rt_hand_nd <dbl> 6.7, 6.2, 6.5, 6.5, 6.5, 6.2, 6.3, 6.3, 6.8, 6.3, 6.2, … $ sqr_rt_hand_nd <dbl> 608920, 247009, 407044, 407895, 425104, 248668, 278080,… $ lg_rt_foot_d <dbl> 6.9, 6.6, 6.8, 6.8, 6.8, 6.5, 6.7, 6.6, 6.6, 6.4, 6.5, … $ sqr_rt_foot_d <dbl> 1018081, 544152, 770884, 814205, 851929, 471511, 605803… $ lg_rt_foot_nd <dbl> 6.9, 6.5, 6.7, 7.2, 6.7, 6.4, 6.6, 6.7, 6.7, 6.7, 6.6, … $ sqr_rt_foot_nd <dbl> 926727, 479325, 617796, 1886960, 648025, 359600, 562000…
~.”mutate(blomkvist_rt, across(rt, round, 0))
# A tibble: 354 × 4
id smoker age rt
<dbl> <chr> <dbl> <dbl>
1 1 former 84 702
2 2 no 37 471
3 3 yes 62 639
4 4 former 85 708
5 5 former 73 607
# … with 349 more rowsmutate(blomkvist_rt, across(rt, ~round(., 0)))
# A tibble: 354 × 4
id smoker age rt
<dbl> <chr> <dbl> <dbl>
1 1 former 84 702
2 2 no 37 471
3 3 yes 62 639
4 4 former 85 708
5 5 former 73 607
# … with 349 more rows~.”~”: we want to make the position of the argument in function explicit..”: the location of the argument.>, ==, ^, +) and when argument is not in first position of supplied function.mutate(blomkvist_rt,
across(rt, ~round(. , 0)), # optional
across(age, ~.^2), # operator
across(rt, ~paste("RT is", .))) # position
# A tibble: 354 × 4
id smoker age rt
<dbl> <chr> <dbl> <chr>
1 1 former 7056 RT is 702
2 2 no 1369 RT is 471
3 3 yes 3844 RT is 639
4 4 former 7225 RT is 708
5 5 former 5329 RT is 607
# … with 349 more rowsc_across()Apply a function across different columns.
Example: get the mean for of all rts for each participant (row).
# Long way mutate(blomkvist, mean_rt = (rt_hand_d + rt_hand_nd + rt_foot_d + rt_foot_nd)/4)
# A tibble: 342 × 8
id age smoker rt_hand_d rt_hand_nd rt_foot_d rt_foot_nd mean_rt
<dbl> <dbl> <chr> <dbl> <dbl> <dbl> <dbl> <dbl>
1 1 84 former 702. 780. 1009 963. 863.
2 2 37 no 471. 497 738. 692. 599.
3 3 62 yes 639. 638 878 786 735.
4 4 85 former 708 639. 902. 1374. 906.
5 5 73 former 607. 652 923 805 747.
# … with 337 more rows
c_across()# Long way mutate(blomkvist, mean_rt = (rt_hand_d + rt_hand_nd + rt_foot_d + rt_foot_nd)/4)
# Still a lot of typing :( blomkvist_rowwise <- rowwise(blomkvist) # each row / ppt is a group mutate(blomkvist_rowwise, mean_rt = mean(c_across(c(rt_hand_d, rt_hand_nd, rt_foot_d, rt_foot_nd))))
# A tibble: 342 × 8
# Rowwise:
id age smoker rt_hand_d rt_hand_nd rt_foot_d rt_foot_nd mean_rt
<dbl> <dbl> <chr> <dbl> <dbl> <dbl> <dbl> <dbl>
1 1 84 former 702. 780. 1009 963. 863.
2 2 37 no 471. 497 738. 692. 599.
3 3 62 yes 639. 638 878 786 735.
4 4 85 former 708 639. 902. 1374. 906.
5 5 73 former 607. 652 923 805 747.
# … with 337 more rows
c_across()# Long way mutate(blomkvist, mean_rt = (rt_hand_d + rt_hand_nd + rt_foot_d + rt_foot_nd)/4)
# That's better :)
blomkvist_rowwise <- rowwise(blomkvist) # each row / ppt is a group
mutate(blomkvist_rowwise, mean_rt = mean(c_across(starts_with("rt_"))))
# A tibble: 342 × 8
# Rowwise:
id age smoker rt_hand_d rt_hand_nd rt_foot_d rt_foot_nd mean_rt
<dbl> <dbl> <chr> <dbl> <dbl> <dbl> <dbl> <dbl>
1 1 84 former 702. 780. 1009 963. 863.
2 2 37 no 471. 497 738. 692. 599.
3 3 62 yes 639. 638 878 786 735.
4 4 85 former 708 639. 902. 1374. 906.
5 5 73 former 607. 652 923 805 747.
# … with 337 more rows
c_across()# Without rowwise grouping returns the grand average:
mutate(blomkvist, mean_rt = mean(c_across(starts_with("rt_"))))
# A tibble: 342 × 8
id age smoker rt_hand_d rt_hand_nd rt_foot_d rt_foot_nd mean_rt
<dbl> <dbl> <chr> <dbl> <dbl> <dbl> <dbl> <dbl>
1 1 84 former 702. 780. 1009 963. 760.
2 2 37 no 471. 497 738. 692. 760.
3 3 62 yes 639. 638 878 786 760.
4 4 85 former 708 639. 902. 1374. 760.
5 5 73 former 607. 652 923 805 760.
# … with 337 more rows
# which is not what we want.
across()filter(blomkvist, rt_hand_d > 1000, rt_hand_nd > 1000, rt_foot_d > 1000, rt_foot_nd > 100)
# A tibble: 12 × 7
id age smoker rt_hand_d rt_hand_nd rt_foot_d rt_foot_nd
<dbl> <dbl> <chr> <dbl> <dbl> <dbl> <dbl>
1 70 85 no 1102. 1352. 1458 1335.
2 96 77 no 1241. 1651. 1217. 1266.
3 127 96 no 1179. 1090 1546 4820
4 152 68 no 1030. 1095 1137 1075.
5 171 92 no 1025. 1026. 1379. 1310.
# … with 7 more rows
across()filter(blomkvist, across(starts_with("rt_"), ~ . > 1000 ))
# A tibble: 12 × 7
id age smoker rt_hand_d rt_hand_nd rt_foot_d rt_foot_nd
<dbl> <dbl> <chr> <dbl> <dbl> <dbl> <dbl>
1 70 85 no 1102. 1352. 1458 1335.
2 96 77 no 1241. 1651. 1217. 1266.
3 127 96 no 1179. 1090 1546 4820
4 152 68 no 1030. 1095 1137 1075.
5 171 92 no 1025. 1026. 1379. 1310.
# … with 7 more rows
across()summarise(blomkvist, rt_hand_d_mean = mean(rt_hand_d),
rt_hand_nd_mean = mean(rt_hand_nd))
# A tibble: 1 × 2
rt_hand_d_mean rt_hand_nd_mean
<dbl> <dbl>
1 638. 640.
across()summarise(blomkvist, across(c(rt_hand_d, rt_hand_nd), mean))
# A tibble: 1 × 2
rt_hand_d rt_hand_nd
<dbl> <dbl>
1 638. 640.
across()summarise(blomkvist, across(starts_with("rt_"), mean))
# A tibble: 1 × 4
rt_hand_d rt_hand_nd rt_foot_d rt_foot_nd
<dbl> <dbl> <dbl> <dbl>
1 638. 640. 893. 871.
across()summarise(blomkvist, across(starts_with("rt_"), mean))
# A tibble: 1 × 4
rt_hand_d rt_hand_nd rt_foot_d rt_foot_nd
<dbl> <dbl> <dbl> <dbl>
1 638. 640. 893. 871.
summarise(blomkvist, across(starts_with("rt_"), list(mean = mean, sd = sd)))
# A tibble: 1 × 8
rt_hand_d_mean rt_hand_d_sd rt_hand_nd_mean rt_hand_nd_sd rt_foot_d_mean
<dbl> <dbl> <dbl> <dbl> <dbl>
1 638. 206. 640. 188. 893.
# … with 3 more variables: rt_foot_d_sd <dbl>, rt_foot_nd_mean <dbl>,
# rt_foot_nd_sd <dbl>
across() to mutate and summarise variablesOpen exercise 6 script.
Called, a long data format.
blomkvist_wide <- select(blomkvist, id, starts_with("rt_"))
blomkvist_wide
# A tibble: 342 × 5
id rt_hand_d rt_hand_nd rt_foot_d rt_foot_nd
<dbl> <dbl> <dbl> <dbl> <dbl>
1 1 702. 780. 1009 963.
2 2 471. 497 738. 692.
3 3 639. 638 878 786
4 4 708 639. 902. 1374.
5 5 607. 652 923 805
# … with 337 more rows
RT is distributed across 4 columns: useful for certain analyses but isn’t tidy.
blomkvist_wide
# A tibble: 342 × 5
id rt_hand_d rt_hand_nd rt_foot_d rt_foot_nd
<dbl> <dbl> <dbl> <dbl> <dbl>
1 1 702. 780. 1009 963.
2 2 471. 497 738. 692.
3 3 639. 638 878 786
4 4 708 639. 902. 1374.
5 5 607. 652 923 805
# … with 337 more rowspivot_longer(blomkvist_wide, cols = starts_with("rt_"))
# A tibble: 1,368 × 3
id name value
<dbl> <chr> <dbl>
1 1 rt_hand_d 702.
2 1 rt_hand_nd 780.
3 1 rt_foot_d 1009
4 1 rt_foot_nd 963.
5 2 rt_hand_d 471.
# … with 1,363 more rowspivot_longer(blomkvist_wide, cols = -id) # shorthand
# A tibble: 1,368 × 3
id name value
<dbl> <chr> <dbl>
1 1 rt_hand_d 702.
2 1 rt_hand_nd 780.
3 1 rt_foot_d 1009
4 1 rt_foot_nd 963.
5 2 rt_hand_d 471.
# … with 1,363 more rowspivot_longer(blomkvist_wide,
cols = -id,
names_to = "variable",
values_to = "rt")
# A tibble: 1,368 × 3
id variable rt
<dbl> <chr> <dbl>
1 1 rt_hand_d 702.
2 1 rt_hand_nd 780.
3 1 rt_foot_d 1009
4 1 rt_foot_nd 963.
5 2 rt_hand_d 471.
# … with 1,363 more rowspivot_longer(blomkvist_wide,
cols = -id,
names_to = "variable",
values_to = "rt")
# A tibble: 1,368 × 3
id variable rt
<dbl> <chr> <dbl>
1 1 rt_hand_d 702.
2 1 rt_hand_nd 780.
3 1 rt_foot_d 1009
4 1 rt_foot_nd 963.
5 2 rt_hand_d 471.
# … with 1,363 more rowspivot_longer(blomkvist_wide,
cols = -id,
names_to = c(".value",
"response_by",
"dominant"),
names_pattern = "(.+)_(.+)_(.+)")
# A tibble: 1,368 × 4
id response_by dominant rt
<dbl> <chr> <chr> <dbl>
1 1 hand d 702.
2 1 hand nd 780.
3 1 foot d 1009
4 1 foot nd 963.
5 2 hand d 471.
# … with 1,363 more rows# Flexible naming pattern
pivot_longer(blomkvist_wide,
cols = -id,
names_to = c(".value",
"response_by",
"dominant"),
names_pattern = "(.+)_(.+)_(.+)")
# A tibble: 1,368 × 4
id response_by dominant rt
<dbl> <chr> <chr> <dbl>
1 1 hand d 702.
2 1 hand nd 780.
3 1 foot d 1009
4 1 foot nd 963.
5 2 hand d 471.
# … with 1,363 more rows# Simplifies to
pivot_longer(blomkvist_wide,
cols = -id,
names_to = c(".value",
"response_by",
"dominant"),
names_sep = "_")
# A tibble: 1,368 × 4
id response_by dominant rt
<dbl> <chr> <chr> <dbl>
1 1 hand d 702.
2 1 hand nd 780.
3 1 foot d 1009
4 1 foot nd 963.
5 2 hand d 471.
# … with 1,363 more rowsbk_long <- pivot_longer(blomkvist_wide,
cols = -id,
names_to = c(".value", "response_by", "dominant"),
names_sep = "_")
bk_long
# A tibble: 1,368 × 4
id response_by dominant rt
<dbl> <chr> <chr> <dbl>
1 1 hand d 702.
2 1 hand nd 780.
3 1 foot d 1009
4 1 foot nd 963.
5 2 hand d 471.
# … with 1,363 more rows
bk_long
# A tibble: 1,368 × 4
id response_by dominant rt
<dbl> <chr> <chr> <dbl>
1 1 hand d 702.
2 1 hand nd 780.
3 1 foot d 1009
4 1 foot nd 963.
5 2 hand d 471.
# … with 1,363 more rowspivot_wider(bk_long,
names_from = response_by,
values_from = rt)
# A tibble: 684 × 4
id dominant hand foot
<dbl> <chr> <dbl> <dbl>
1 1 d 702. 1009
2 1 nd 780. 963.
3 2 d 471. 738.
4 2 nd 497 692.
5 3 d 639. 878
# … with 679 more rowsbk_long
# A tibble: 1,368 × 4
id response_by dominant rt
<dbl> <chr> <chr> <dbl>
1 1 hand d 702.
2 1 hand nd 780.
3 1 foot d 1009
4 1 foot nd 963.
5 2 hand d 471.
# … with 1,363 more rowspivot_wider(bk_long,
names_from = c(response_by, dominant),
values_from = rt)
# A tibble: 342 × 5
id hand_d hand_nd foot_d foot_nd
<dbl> <dbl> <dbl> <dbl> <dbl>
1 1 702. 780. 1009 963.
2 2 471. 497 738. 692.
3 3 639. 638 878 786
4 4 708 639. 902. 1374.
5 5 607. 652 923 805
# … with 337 more rowsbk_long
# A tibble: 1,368 × 4
id response_by dominant rt
<dbl> <chr> <chr> <dbl>
1 1 hand d 702.
2 1 hand nd 780.
3 1 foot d 1009
4 1 foot nd 963.
5 2 hand d 471.
# … with 1,363 more rowspivot_wider(bk_long,
names_from = c(response_by, dominant),
values_from = rt,
names_prefix = "rt_")
# A tibble: 342 × 5
id rt_hand_d rt_hand_nd rt_foot_d rt_foot_nd
<dbl> <dbl> <dbl> <dbl> <dbl>
1 1 702. 780. 1009 963.
2 2 471. 497 738. 692.
3 3 639. 638 878 786
4 4 708 639. 902. 1374.
5 5 607. 652 923 805
# … with 337 more rowsContinue with exercise 7
# Combine two datasets side-by-side bind_cols(data_1, data_2) # Stacking two data sets bind_rows(data_1, data_2) # Keep all rows of data_1 and add data_2 left_join(data_1, data_2) # Keep all rows of data_2 and add data_1 right_join(data_1, data_2) # Include all rows of both data sets full_join(data_1, data_2) # Include data that is present in both data sets inner_join(data_1, data_2)
bind_cols()bk_id_age <- select(blomkvist, id, age) bk_id_age
# A tibble: 354 × 2
id age
<dbl> <dbl>
1 1 84
2 2 37
3 3 62
4 4 85
5 5 73
# … with 349 more rowsbk_med_smoke <- select(blomkvist, medicine, smoker) bk_med_smoke
# A tibble: 354 × 2
medicine smoker
<dbl> <chr>
1 8 former
2 1 no
3 0 yes
4 4 former
5 5 former
# … with 349 more rowsbind_cols()# Combine two datasets side-by-side bind_cols(bk_id_age, bk_med_smoke)
# A tibble: 354 × 4
id age medicine smoker
<dbl> <dbl> <dbl> <chr>
1 1 84 8 former
2 2 37 1 no
3 3 62 0 yes
4 4 85 4 former
5 5 73 5 former
# … with 349 more rows
bind_rows()bk_former_smokers <- filter(blomkvist, smoker == "former") bk_smokers <- filter(blomkvist, smoker == "yes")
bind_rows()# Stacking two data sets bk_smoking <- bind_rows(bk_former_smokers, bk_smokers)
slice_head(bk_smoking, n = 1)
# A tibble: 1 × 9
id sex age medicine smoker rt_hand_d rt_hand_nd rt_foot_d rt_foot_nd
<dbl> <chr> <dbl> <dbl> <chr> <dbl> <dbl> <dbl> <dbl>
1 1 male 84 8 former 702. 780. 1009 963.
slice_tail(bk_smoking, n = 1)
# A tibble: 1 × 9
id sex age medicine smoker rt_hand_d rt_hand_nd rt_foot_d rt_foot_nd
<dbl> <chr> <dbl> <dbl> <chr> <dbl> <dbl> <dbl> <dbl>
1 342 male 60 3 yes 708 686 909 854
This also works when columns are not in the same order as long as the names and types match.
bk_smoker <- select(blomkvist, smoker, age) bk_smoker_o40 <- filter(bk_smoker, age > 40)
bk_sex <- select(blomkvist, sex, age) bk_sex_u50 <- filter(bk_sex, age < 50)
bk_smoker_o40
# A tibble: 248 × 2 smoker age <chr> <dbl> 1 former 84 2 yes 62 3 former 85 4 former 73 5 no 65 # … with 243 more rows
bk_sex_u50
# A tibble: 144 × 2 sex age <chr> <dbl> 1 female 37 2 female 30 3 female 49 4 female 25 5 male 27 # … with 139 more rows
left_join()# Keep all rows of bk_sex_u50
bk_joined <- left_join(bk_sex_u50,
bk_smoker_o40, by = "age")
bk_joined
# A tibble: 290 × 3 sex age smoker <chr> <dbl> <chr> 1 female 37 <NA> 2 female 30 <NA> 3 female 49 former 4 female 49 no 5 female 49 former # … with 285 more rows
range(bk_joined$age)
[1] 20 49
right_join()# Keep all rows of bk_smoker_o40
bk_joined <- right_join(bk_sex_u50,
bk_smoker_o40, by = "age")
bk_joined
# A tibble: 394 × 3 sex age smoker <chr> <dbl> <chr> 1 female 49 former 2 female 49 no 3 female 49 former 4 female 49 former 5 female 49 no # … with 389 more rows
range(bk_joined$age)
[1] 41 99
full_join()# Include all rows of both data sets
bk_joined <- full_join(bk_sex_u50,
bk_smoker_o40, by = "age")
bk_joined
# A tibble: 500 × 3 sex age smoker <chr> <dbl> <chr> 1 female 37 <NA> 2 female 30 <NA> 3 female 49 former 4 female 49 no 5 female 49 former # … with 495 more rows
range(bk_joined$age)
[1] 20 99
inner_join()# Include only info that is present in both data sets
bk_joined <- inner_join(bk_sex_u50,
bk_smoker_o40, by = "age")
bk_joined
# A tibble: 184 × 3 sex age smoker <chr> <dbl> <chr> 1 female 49 former 2 female 49 no 3 female 49 former 4 female 49 former 5 female 49 no # … with 179 more rows
range(bk_joined$age)
[1] 41 49
Continue with exercise 8
%>%%>% moves or “pipes” the result forward into the next functionf(x) is the same as x %>% f()Ctrl + Shift + Mselect(data, myvar1, myvar2) # or data %>% select(myvar1, myvar2)
*assumes first argument is data
%>%# Instead of
data_1 <- first_step(data)
data_2 <- second_step(data_2)
data_3 <- third_step(data_3)
data_4 <- fourth_step(data_4)
# or
fourth_step(
third_step(
second_step(
first_step(data)
)
)
)# Just do data %>% first_step() %>% second_step() %>% third_step() %>% fourth_step()
%>%bk <- read_csv("../data/blomkvist.csv")
bk_rts <- select(bk, id, starts_with("rt_"))
bk_rts_flt <- filter(bk_rts, rt_hand_d > 1500)
bk_rts_flt
# A tibble: 4 × 5
id rt_hand_d rt_hand_nd rt_foot_d rt_foot_nd
<dbl> <dbl> <dbl> <dbl> <dbl>
1 181 1552. 1072. 1068 1196.
2 262 1595. 1162. 1508 1496.
3 290 2076 1845 17094. 3874.
4 305 1527 1416. 1713. 2471.read_csv("../data/blomkvist.csv") %>%
select(id, starts_with("rt_")) %>%
filter(rt_hand_d > 1500)
# A tibble: 4 × 5
id rt_hand_d rt_hand_nd rt_foot_d rt_foot_nd
<dbl> <dbl> <dbl> <dbl> <dbl>
1 181 1552. 1072. 1068 1196.
2 262 1595. 1162. 1508 1496.
3 290 2076 1845 17094. 3874.
4 305 1527 1416. 1713. 2471.%>%Continue with exercise 9
Andrews, Mark, and Jens Roeser. 2021. Psyntur: Helper Tools for Teaching Statistical Data Analysis. https://CRAN.R-project.org/package=psyntur.
Blomkvist, Andreas W., Fredrik Eika, Martin T. Rahbek, Karin D. Eikhof, Mette D. Hansen, Malene Søndergaard, Jesper Ryg, Stig Andersen, and Martin G. Jørgensen. 2017. “Reference Data on Reaction Time and Aging Using the Nintendo Wii Balance Board: A Cross-Sectional Study of 354 Subjects from 20 to 99 Years of Age.” PLoS One 12 (12): e0189598. https://doi.org/10.1371/journal.pone.0189598.
Roeser, Jens, Sven De Maeyer, Mariëlle Leijten, and Luuk Van Waes. 2021. “Modelling Typing Disfluencies as Finite Mixture Process.” Reading and Writing, 1–26.