Tidyverse

From 太極
Jump to navigation Jump to search

Tidyverse

   Import
     |
     | readr, readxl
     | haven, DBI, httr   +----- Visualize ------+
     |                    |    ggplot2, ggvis    |
     |                    |                      |
   Tidy ------------- Transform 
   tibble               dplyr                   Model 
   tidyr                  |                    broom
                          +------ Model ---------+

Cheat sheet

The cheat sheets are downloaded from RStudio

Online

Animation to explain

tidyexplain - Tidy Animated Verbs

Examples

A Gentle Introduction to Tidy Statistics in R

A Gentle Introduction to Tidy Statistics in R by Thomas Mock on RStudio webinar. Good coverage with step-by-step explanation. See part 1 & part 2 about the data and markdown document. All documents are available in github repository.

Task R code Graph
Load the libraries 
library(tidyverse)
library(readxl)
library(broom)
library(knitr)
Read Excel file 
raw_df <- readxl::read_xlsx("ad_treatment.xlsx")

dplyr::glimpse(raw_df)
Check distribution 
g2 <- ggplot(raw_df, aes(x = age)) +
  geom_density(fill = "blue")
g2
raw_df %>% summarize(min = min(age),
                     max = max(age))
File:Check dist.svg
Data cleaning 
raw_df %>% 
  summarize(na_count = sum(is.na(mmse)))
Experimental variables

levels 

# check Ns and levels for our variables
table(raw_df$drug_treatment, raw_df$health_status)
table(raw_df$drug_treatment, raw_df$health_status, raw_df$sex)

# tidy way of looking at variables
raw_df %>% 
  group_by(drug_treatment, health_status, sex) %>% 
  count()
Visual Exploratory

Data Analysis 

ggplot(data = raw_df, # add the data
       aes(x = drug_treatment, y = mmse, # set x, y coordinates
           color = drug_treatment)) +    # color by treatment
  geom_boxplot() +
  facet_grid(~health_status)
File:Onefacet.svg
Summary Statistics 
raw_df %>% 
  glimpse()
sum_df <- raw_df %>% 
            mutate(
              sex = factor(sex, 
                  labels = c("Male", "Female")),
              drug_treatment =  factor(drug_treatment, 
                  levels = c("Placebo", "Low dose", "High Dose")),
              health_status = factor(health_status, 
                  levels = c("Healthy", "Alzheimer's"))
              ) %>% 
            group_by(sex, health_status, drug_treatment # group by categorical variables
              ) %>%  
            summarize(
              mmse_mean = mean(mmse),      # calc mean
              mmse_se = sd(mmse)/sqrt(n()) # calc standard error
              ) %>%  
            ungroup() # ungrouping variable is a good habit to prevent errors

kable(sum_df)

write.csv(sum_df, "adx37_sum_stats.csv")
Plotting summary

statistics 

g <- ggplot(data = sum_df, # add the data
       aes(x = drug_treatment,  #set x, y coordinates
           y = mmse_mean,
           group = drug_treatment,  # group by treatment
           color = drug_treatment)) +    # color by treatment
  geom_point(size = 3) + # set size of the dots
  facet_grid(sex~health_status) # create facets by sex and status
g
File:Twofacets.svg
ANOVA 
# set up the statistics df
stats_df <- raw_df %>% # start with data
   mutate(drug_treatment = factor(drug_treatment, levels = c("Placebo", "Low dose", "High Dose")),
         sex = factor(sex, labels = c("Male", "Female")),
         health_status = factor(health_status, levels = c("Healthy", "Alzheimer's")))

glimpse(stats_df)
# this gives main effects AND interactions
ad_aov <- aov(mmse ~ sex * drug_treatment * health_status, 
        data = stats_df)

summary(ad_aov)


# this extracts ANOVA output into a nice tidy dataframe
tidy_ad_aov <- tidy(ad_aov)
# which we can save to Excel
write.csv(tidy_ad_aov, "ad_aov.csv")
Post-hocs 
# pairwise t.tests
ad_pairwise <- pairwise.t.test(stats_df$mmse,
                               stats_df$sex:stats_df$drug_treatment:stats_df$health_status, 
                               p.adj = "none")
# look at the posthoc p.values in a tidy dataframe
kable(head(tidy(ad_pairwise)))


# call and tidy the tukey posthoc
tidy_ad_tukey <- tidy(
                      TukeyHSD(ad_aov, 
                              which = 'sex:drug_treatment:health_status'))
Publication plot 
sig_df <- tribble(
  ~drug_treatment, ~ health_status, ~sex, ~mmse_mean,
  "Low dose", "Alzheimer's", "Male", 17,
  "High Dose", "Alzheimer's", "Male", 25,
  "Low dose", "Alzheimer's", "Female", 18, 
  "High Dose", "Alzheimer's", "Female", 24
  )

sig_df <- sig_df %>% 
  mutate(drug_treatment = factor(drug_treatment, levels = c("Placebo", "Low dose", "High Dose")),
         sex = factor(sex, levels = c("Male", "Female")),
         health_status = factor(health_status, levels = c("Healthy", "Alzheimer's")))
sig_df
# plot of cognitive function health and drug treatment
g1 <- ggplot(data = sum_df, 
       aes(x = drug_treatment, y = mmse_mean, fill = drug_treatment,  
           group = drug_treatment)) +
  geom_errorbar(aes(ymin = mmse_mean - mmse_se, 
                    ymax = mmse_mean + mmse_se), width = 0.5) +
  geom_bar(color = "black", stat = "identity", width = 0.7) +
  
  facet_grid(sex~health_status) +
  theme_bw() +
  scale_fill_manual(values = c("white", "grey", "black")) +
  theme(legend.position = "NULL",
        legend.title = element_blank(),
        axis.title = element_text(size = 20),
        legend.background = element_blank(),
        panel.grid.major = element_blank(), 
        panel.grid.minor = element_blank(),
        axis.text = element_text(size = 12)) +
  geom_text(data = sig_df, label = "*", size = 8) +
  labs(x = "\nDrug Treatment", 
       y = "Cognitive Function (MMSE)\n",
       caption = "\nFigure 1. Effect of novel drug treatment AD-x37 on cognitive function in 
                    healthy and demented elderly adults. 
                  \nn = 100/treatment group (total n = 600), * indicates significance 
                    at p < 0.001")
g1

# save the graph!
ggsave("ad_publication_graph.png", g1, height = 7, width = 8, units = "in")
File:Ad public.svg

Opioid prescribing habits in texas

https://juliasilge.com/blog/texas-opioids/.

  • It can read multiple sheets (27 sheets) at a time and merge them by rows.
  • case_when(): A general vectorised if. This function allows you to vectorise multiple if_else() statements. 
    x %>% mutate(group = case_when(
  PredScore > quantile(PredScore, .5) ~ 'High',
  PredScore < quantile(PredScore, .5) ~ 'Low',
  TRUE ~ NA_character_
))

Useful dplyr functions (with examples)

https://sw23993.wordpress.com/2017/07/10/useful-dplyr-functions-wexamples/

Supervised machine learning case studies in R

Supervised machine learning case studies in R - A Free, Interactive Course Using Tidy Tools.

Time series data

Calculating change from baseline

group_by() + mutate() + ungroup(). We can accomplish the task by using split() + lapply() + do.call(). 

trial_data_chg <- trial_data %>%
  arrange(USUBJID, AVISITN) %>%
  group_by(USUBJID) %>%
  mutate(CHG = AVAL - AVAL[1L]) %>%
  ungroup()

# If the baseline is missing
trial_data_chg2 <- trial_data %>%
  group_by(USUBJID) %>%
  mutate(
    CHG = if (any(AVISIT == "Baseline")) AVAL - AVAL[AVISIT == "Baseline"] else NA
  ) %>%
  ungroup()

Show all possible group combinations

Install on Ubuntu

sudo apt install r-cran-tidyverse

# Ubuntu >= 18.04. However, I get unmet dependencies errors on R 3.5.3.
# r-cran-curl : Depends: r-api-3.4
sudo apt-get install r-cran-curl r-cran-openssl r-cran-xml2

# Works fine on Ubuntu 16.04, 18.04, 20.04
sudo apt install libcurl4-openssl-dev libssl-dev libxml2-dev

80 R packages will be installed after tidyverse has been installed.

For RStudio server docker version (Debian 10), I also need to install zlib1g-dev

Install on Raspberry Pi/(ARM based) Chromebook

In additional to the requirements of installing on Ubuntu, I got an error when it is installing a dependent package fs: undefined symbol: pthread_atfork. The fs package version is 1.2.6. The solution is to add one line in fs/src/Makevars file and then install the "fs" package using the source on the local machine.

5 most useful data manipulation functions

  • subset() for making subsets of data (natch)
  • merge() for combining data sets in a smart and easy way
  • melt()-reshape2 package for converting from wide to long data formats. See an example here where we want to combine multiple columns of values into 1 column. melt() is replaced by gather().
  • dcast()-reshape2 package for converting from long to wide data formats (or just use tapply()), and for making summary tables
  • ddply()-plyr package for doing split-apply-combine operations, which covers a huge swath of the most tricky data operations

Miscellaneous examples using tibble or dplyr packages

Move a column to rownames

?tibble::column_to_rownames 

# It assumes the input data frame has no row names; otherwise we will get
# Error: `df` must be a data frame without row names in `column_to_rownames()`
# 
tibble::column_to_rownames(data.frame(x=letters[1:5], y = rnorm(5)), "x")

Move rownames to a variable

https://tibble.tidyverse.org/reference/rownames.html 

tibble::rownames_to_column(trees, "newVar")
# Still a data frame

Old way add_rownames() 

data.frame(x=1:5, y=2:6) %>% magrittr::set_rownames(letters[1:5]) %>% add_rownames("newvar")
# tibble object

Rename variables

dplyr::rename(os, newName = oldName)

Drop a variable

select(df, -x)

Drop a level

group_by() has a .drop argument so you can also group by factor levels that don't appear in the data. See this example.

Remove rownames

tibble has_rownames(), rownames_to_column(), column_to_rownames() 

has_rownames(mtcars)
#> [1] TRUE

# Remove row names
remove_rownames(mtcars) %>% has_rownames()
#> [1] FALSE

> tibble::has_rownames(trees)
[1] FALSE
> rownames(trees)
 [1] "1"  "2"  "3"  "4"  "5"  "6"  "7"  "8"  "9"  "10" "11" "12" "13" "14" "15"
[16] "16" "17" "18" "19" "20" "21" "22" "23" "24" "25" "26" "27" "28" "29" "30"
[31] "31"
> rownames(trees) <- NULL
> rownames(trees)
 [1] "1"  "2"  "3"  "4"  "5"  "6"  "7"  "8"  "9"  "10" "11" "12" "13" "14" "15"
[16] "16" "17" "18" "19" "20" "21" "22" "23" "24" "25" "26" "27" "28" "29" "30"
[31] "31"

relocate: change column order

relocate() 

# Move Petal.Width column to appear next to Sepal.Width
iris %>% relocate(Petal.Width, .after = Sepal.Width) %>% head() 

# Move Petal.Width to the last column
iris %>% relocate(Petal.Width, .after = last_col()) %>% head()

pull: extract a single column

pull() 

x <- iris %>% filter(Species == 'setosa') %>% select(Sepal.Length) %>% pull()
y <- iris %>% filter(Species == 'virginica') %>% select(Sepal.Length) %>% pull()
t.test(x, y)

Anonymous functions

data.table

Fast aggregation of large data (e.g. 100GB in RAM or just several GB size file), fast ordered joins, fast add/modify/delete of columns by group using no copies at all, list columns and a fast file reader (fread).

Note: data.table has its own ways (cf base R and dplyr) to subset columns.

Some resources:

OpenMP enabled compiler for Mac. This instruction works on my Mac El Capitan (10.11.6) when I need to upgrade the data.table version from 1.11.4 to 1.11.6.

Question: how to make use multicore with data.table package?

dtplyr

https://www.tidyverse.org/blog/2019/11/dtplyr-1-0-0/

reshape & reshape2 (superceded by tidyr package)

tidyr

Missing values

Handling Missing Values in R using tidyr

Pivot

  • A Tidy Transcriptomics introduction to RNA-Seq analyses 
    data %>% pivot_longer(cols = c("counts", "counts_scaled"), names_to = "source", values_to = "abundance")
  • Using R: setting a colour scheme in ggplot2. Note the new (default) column names value and name after the function pivot_longer(data, cols). 
    set(1)
dat1 <- data.frame(y=rnorm(10), x1=rnorm(10), x2=rnorm(10))
dat2 <- pivot_longer(dat1, -y)
head(dat2, 2)
# A tibble: 2 x 3
      y name   value
  <dbl> <chr>  <dbl>
1 -1.28 x1     0.717
2 -1.28 x2    -0.320

dat3 <- pivot_wider(dat2)
range(dat1 - dat3)

unnest()

Benchmark

An evolution of reshape2. It's designed specifically for data tidying (not general reshaping or aggregating) and works well with dplyr data pipelines.

Make wide tables long with gather() (see 6.3.1 of Efficient R Programming) 

library(tidyr)
library(efficient)
data(pew) # wide table
dim(pew) # 18 x 10,  (religion, '<$10k', '$10--20k', '$20--30k', ..., '>150k') 
pewt <- gather(data = pew, key = Income, value = Count, -religion)
dim(pew) # 162 x 3,  (religion, Income, Count)

args(gather)
# function(data, key, value, ..., na.rm = FALSE, convert = FALSE, factor_key = FALSE)

where the three arguments of gather() requires:

  • data: a data frame in which column names will become row values. If the data is a matrix, use %>% as.data.frame() beforehand.
  • key: the name of the categorical variable into which the column names in the original datasets are converted.
  • value: the name of cell value columns

In this example, the 'religion' column will not be included (-religion).

dplyr, plyr packages

  • plyr package suffered from being slow in some cases. dplyr addresses this by porting much of the computation to C++. Another additional feature is the ability to work with data stored directly in an external database. The benefits of doing this are the data can be managed natively in a relational database, queries can be conducted on that database, and only the results of query returned.
  • Essential functions: 3 rows functions, 3 column functions and 1 mixed function.
           select, mutate, rename, recode
            +------------------+
filter      +                  +
arrange     +                  +
group_by    +                  +
drop_na     +                  +
ungroup     + summarise        +
            +------------------+
  • These functions works on data frames and tibble objects. Note stats package also has a filter() function for time series data. If we have not loaded the dplyr package, the filter() function below will give an error (count() also is from dplyr).
iris %>% filter(Species == "setosa") %>% count()
head(iris %>% filter(Species == "setosa") %>% arrange(Sepal.Length))
  • dplyr tutorial from PH525x series (Biomedical Data Science by Rafael Irizarry and Michael Love). For select() function, some additional options to select columns based on a specific criteria include
    • start_with()/ ends_with() = Select columns that start/end with a character string
    • contains() = Select columns that contain a character string
    • matches() = Select columns that match a regular expression
    • one_of() = Select columns names that are from a group of names
  • Data Transformation in the book R for Data Science. Five key functions in the dplyr package: 
# filter
jan1 <- filter(flights, month == 1, day == 1)
filter(flights, month == 11 | month == 12)
filter(flights, arr_delay <= 120, dep_delay <= 120)
df <- tibble(x = c(1, NA, 3))
filter(df, x > 1)
filter(df, is.na(x) | x > 1)

# arrange
arrange(flights, year, month, day)
arrange(flights, desc(arr_delay))

# select
select(flights, year, month, day)
select(flights, year:day)
select(flights, -(year:day))

# mutate
flights_sml <- select(flights, 
  year:day, 
  ends_with("delay"), 
  distance, 
  air_time
)
mutate(flights_sml,
  gain = arr_delay - dep_delay,
  speed = distance / air_time * 60
)
# if you only want to keep the new variables
transmute(flights,
  gain = arr_delay - dep_delay,
  hours = air_time / 60,
  gain_per_hour = gain / hours
)

# summarise()
by_day <- group_by(flights, year, month, day)
summarise(by_day, delay = mean(dep_delay, na.rm = TRUE))

# pipe. Note summarise() can return more than 1 variable.
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")
flights %>% 
  group_by(year, month, day) %>% 
  summarise(mean = mean(dep_delay, na.rm = TRUE))

select()

Select columns from a data frame

select() + everything()

If we want one particular column (say the dependent variable y) to appear first or last in the dataset. We can use the everything(). 

iris %>% select(Species, everything()) %>% head()
iris %>% select(-Species, everything()) %>% head() # put Species to the last col

plyr::rbind.fill()

Videos

dbplyr

https://dbplyr.tidyverse.org/articles/dbplyr.html

stringr

  • stringr is part of the tidyverse but is not a core package. You need to load it separately.
  • https://www.rstudio.com/wp-content/uploads/2016/09/RegExCheatsheet.pdf
  • stringr Cheat sheet (2 pages, this will immediately download the pdf file)
    • Detect Matches: str_detect(), str_which(), str_count(), str_locate()
    • Subset: str_sub(), str_subset(), str_extract(), str_match()
    • Manage Lengths: str_length(), str_pad(), str_trunc(), str_trim()
    • Mutate Strings: str_sub(), str_replace(), str_replace_all(), str_remove()
      • Case Conversion: str_to_lower(), str_to_upper(), str_to_title()
    • Joint and Split: str_c(), str_dup(), str_split_fixed(), str_glue(), str_glue_date()
  • Efficient data carpentry → Regular expressions from Efficient R programming book by Gillespie & Lovelace.

magrittr

x %>% f     # f(x)
x %>% f(y)  # f(x, y)
x %>% f(arg=y)  # f(x, arg=y)
x %>% f(z, .) # f(z, x)
x %>% f(y) %>% g(z)  #  g(f(x, y), z)

x %>% select(which(colSums(!is.na(.))>0))  # remove columns with all missing data
x %>% select(which(colSums(!is.na(.))>0)) %>% filter((rowSums(!is.na(.))>0)) # remove all-NA columns _and_ rows
suppressPackageStartupMessages(library("dplyr"))
starwars %>%
  filter(., height > 200) %>%
  select(., height, mass) %>%
  head(.)
# instead of 
starwars %>%
  filter(height > 200) %>%
  select(height, mass) %>%
  head
iris$Species
iris[["Species"]]

iris %>%
`[[`("Species")

iris %>%
`[[`(5)

iris %>%
  subset(select = "Species")
  • Split-apply-combine: group + summarize + sort/arrange + top n. The following example is from Efficient R programming.
data(wb_ineq, package = "efficient")
wb_ineq %>% 
  filter(grepl("g", Country)) %>%
  group_by(Year) %>%
  summarise(gini = mean(gini, na.rm  = TRUE)) %>%
  arrange(desc(gini)) %>%
  top_n(n = 5)
f <- function(x) {
  (y - x) %>% 
    '^'(2) %>% 
    sum %>%
    '/'(length(x)) %>% 
    sqrt %>% 
    round(2)
}
# Examples from R for Data Science-Import, Tidy, Transform, Visualize, and Model
diamonds <- ggplot2::diamonds
diamonds2 <- diamonds %>% dplyr::mutate(price_per_carat = price / carat)

pryr::object_size(diamonds)
pryr::object_size(diamonds2)
pryr::object_size(diamonds, diamonds2)

rnorm(100) %>% matrix(ncol = 2) %>% plot() %>% str()
rnorm(100) %>% matrix(ncol = 2) %T>% plot() %>% str() # 'tee' pipe
    # %T>% works like %>% except that it returns the lefthand side (rnorm(100) %>% matrix(ncol = 2))  
    # instead of the righthand side.

# If a function does not have a data frame based api, you can use %$%.
# It explodes out the variables in a data frame.
mtcars %$% cor(disp, mpg) 

# For assignment, magrittr provides the %<>% operator
mtcars <- mtcars %>% transform(cyl = cyl * 2) # can be simplified by
mtcars %<>% transform(cyl = cyl * 2)

Upsides of using magrittr: no need to create intermediate objects, code is easy to read.

When not to use the pipe

  • your pipes are longer than (say) 10 steps
  • you have multiple inputs or outputs
  • Functions that use the current environment: assign(), get(), load()
  • Functions that use lazy evaluation: tryCatch(), try()

Dollar sign .$

%$%

Expose the names in lhs to the rhs expression. This is useful when functions do not have a built-in data argument. 

lhs %$% rhs
# lhs:	A list, environment, or a data.frame.
# rhs: An expression where the names in lhs is available.

iris %>%
  subset(Sepal.Length > mean(Sepal.Length)) %$%
  cor(Sepal.Length, Sepal.Width)

set_rownames() and set_colnames()

https://stackoverflow.com/a/56613460, https://www.rdocumentation.org/packages/magrittr/versions/1.5/topics/extract 

data.frame(x=1:5, y=2:6) %>% magrittr::set_rownames(letters[1:5])

cbind(1:5, 2:6) %>% magrittr::set_colnames(letters[1:2])

purrr: : Functional Programming Tools

  • Getting started with the purrr package in R, especially the map() function. 
    library(purrr)
# map() is a replacement of lapply()
# lapply(dat, function(x) mean(x$Open))
map(dat, function(x)mean(x$Open))  

# map allows us to bypass the function function. 
# Using a tilda (~) in place of function and a dot (.) in place of x
map(dat, ~mean(.$Open))

# map allows you to specify the structure of your output.
map_dbl(dat, ~mean(.$Open))

# map2() is a replacement of mapply()
# mapply(function(x,y)plot(x$Close, type = "l", main = y), x = dat, y = stocks)
map2(dat, stocks, ~plot(.x$Close, type="l", main = .y))

data <- map(paths, read.csv)
data <- map_dfr(paths, read.csv, id = "path")

out1 <- mtcars %>% map_dbl(mean, na.rm = TRUE)
out2 <- mtcars %>% map_dbl(median, na.rm = TRUE)

forcats

https://forcats.tidyverse.org/

JAMA retraction after miscoding – new Finalfit function to check recoding

outer()

Genomic sequence

  • chartr
> yourSeq <- "AAAACCCGGGTTTNNN"
> chartr("ACGT", "TGCA", yourSeq)
[1] "TTTTGGGCCCAAANNN"

broom

broom: Convert Statistical Analysis Objects into Tidy Tibbles

Especially the tidy() function. 

R> str(survfit(Surv(time, status) ~ x, data = aml))
List of 17
 $ n        : int [1:2] 11 12
 $ time     : num [1:20] 9 13 18 23 28 31 34 45 48 161 ...
 $ n.risk   : num [1:20] 11 10 8 7 6 5 4 3 2 1 ...
 $ n.event  : num [1:20] 1 1 1 1 0 1 1 0 1 0 ...
...

R> tidy(survfit(Surv(time, status) ~ x, data = aml))
# A tibble: 20 x 9
    time n.risk n.event n.censor estimate std.error conf.high conf.low strata         
   <dbl>  <dbl>   <dbl>    <dbl>    <dbl>     <dbl>     <dbl>    <dbl> <chr>          
 1     9     11       1        0   0.909     0.0953     1       0.754  x=Maintained   
 2    13     10       1        1   0.818     0.142      1       0.619  x=Maintained   
...
18    33      3       1        0   0.194     0.627      0.664   0.0569 x=Nonmaintained
19    43      2       1        0   0.0972    0.945      0.620   0.0153 x=Nonmaintained
20    45      1       1        0   0       Inf         NA      NA      x=Nonmaintained

lobstr package - dig into the internal representation and structure of R objects

lobstr 1.0.0

tidymodels

Other packages

janitor

How to Clean Data: {janitor} Package

funneljoin

Retrieved from "https://wiki.taichimd.us/index.php?title=Tidyverse&oldid=26793"