2. Tidyverse worflow

options(warn=-1)
options(messages=-1)
options(tidyverse.quiet = TRUE)
options(dplyr.summarise.inform = FALSE)
shhh <- suppressWarnings # It's a library, so shhh!
shhh(library(tidyverse))

## ---------------------------------------------------------------------------------------------------
vow.dur <- read.table("https://bit.ly/2Iw7kn7", header=TRUE, sep="\t")

# main dplyr functions
vow.dur %>%
  filter(context == "voiceless" & Vow_dur_ms > mean(Vow_dur_ms, na.rm = TRUE)) %>%
  arrange(desc(Vow_dur_ms)) %>%
#   select(contains("_")) %>%
# explain why some vow_dur > mean
  mutate(mean_vow_dur = mean(Vow_dur_ms, na.rm = TRUE)) %>%
  rename(vowel_duration = Vow_dur_ms)

A data.frame: 7 × 6

vowel	US_state	position	context	vowel_duration	mean_vow_dur
<fct>	<fct>	<fct>	<fct>	<dbl>	<dbl>
aɪ	NorthC	Ccontext	voiceless	216.6	203.9143
æ	NorthC	Ccontext	voiceless	214.8	203.9143
aɪ	NorthC	Vemph	voiceless	209.2	203.9143
æ	NorthC	Vemph	voiceless	202.7	203.9143
æ	Ohio	Ccontext	voiceless	198.5	203.9143
e	NorthC	Ccontext	voiceless	193.9	203.9143
e	NorthC	Vemph	voiceless	191.7	203.9143

# advanced dplyr functions

vow.dur %>%
  group_by(US_state, context) %>%
  summarise(mean = mean(Vow_dur_ms))

A grouped_df: 6 × 3

US_state	context	mean
<fct>	<fct>	<dbl>
NorthC	voiced	235.74
NorthC	voiceless	186.26
Ohio	voiced	213.90
Ohio	voiceless	156.94
Wisc	voiced	194.59
Wisc	voiceless	147.14