- Invocation and the Read-Eval-Print Loop
- Constants
- Parsing of R code
- Strings and names
- Functions and operators
- Comments
- Working directory
- Workspace
- Search path
- Documentation
Recap
3rd-Party Libraries (Packages)
- Compilation of functions and/or data
- Made to ease specific (repetitive) tasks
- Some are general-purpose, others domain-specific
- Currently, 17678 packages on CRAN
- Other locations - Bioconductor, GitHub
- See CRAN Task Views
A hypothetical package called pkg has a function, pause():
- To install from CRAN -
install.packages("pkg") - To learn about the package -
help(package = "pkg") - To use
pkgin R session -library(pkg)(orrequire) - Function
pause()can be used one-off -pkg::pause() - To detach
pkgfrom search path -detach("package:pkg") - To uninstall
pkguseremove.packages("pkg")
Importing Data
- Note: There are inbuilt datasets
- Import from MS Excel using the readxl package
- This object is called a data frame
- Has observations (rows) and variables (columns)
dat <- readxl::read_excel("whostat2005_coverage.xls")
# Alternatively:
# library(readxl)
# dat <- read_excel("whostat2005_coverage.xls")
Data exploration
View()(to find out about it -?View)
Data exploration
dim()head()ortail()colnames()str()
dim(dat)
## [1] 216 21
head(dat)
## ...1 ...2 ...3 ...4 ...5 ## 1 NA NA <NA> WHO NA ## 2 NA NA Country region NA ## 3 NA NA <NA> <NA> NA ## 4 NA NA <NA> <NA> NA ## 5 NA NA <NA> <NA> NA ## 6 1 NA Afghanistan EMR NA ## Immunization coverage (%) among 1-year-oldsa ...7 ...8 Antenatal ...10 ## 1 Measles DTP3 HepB3 care coverageb <NA> ## 2 <NA> <NA> <NA> (%) year ## 3 <NA> <NA> <NA> <NA> <NA> ## 4 2003 2003 2003 <NA> <NA> ## 5 <NA> <NA> <NA> <NA> <NA> ## 6 50 54 0 52 2003 ## Births attended by ...12 Contraceptive ...14 ## 1 skilled health personnelb <NA> prevalence rateb <NA> ## 2 (%) year (%) year ## 3 <NA> <NA> <NA> <NA> ## 4 <NA> <NA> <NA> <NA> ## 5 <NA> <NA> <NA> <NA> ## 6 14 2003 4 2000 ## Children under-5 using ...16 TB detection ...18 TB treatment ## 1 insecticide-treated netsc <NA> rate under DOTSd <NA> success under DOTSd ## 2 (%) year (%) year (%) ## 3 <NA> <NA> <NA> <NA> <NA> ## 4 <NA> <NA> <NA> <NA> <NA> ## 5 <NA> <NA> <NA> <NA> <NA> ## 6 ... ... 18 2003 87 ## ...20 Antiretroviral ## 1 <NA> therapy coveragee ## 2 year (%) ## 3 <NA> <NA> ## 4 <NA> Dec 2004 ## 5 <NA> <NA> ## 6 2002 ...
tail(dat)
## ...1 ...2 ## 211 NA NA ## 212 NA NA ## 213 NA NA ## 214 NA NA ## 215 NA NA ## 216 NA NA ## ...3 ## 211 … data not available or not applicable. ## 212 aWorld Health Organization, Department of Immunization Vaccines and Biologicals, Vaccine Assessment and Monitoring Team. (http//www.who.int/vaccines-surveillance, accessed on 16 April 2005) ## 213 bThe World Health Report 2005: make every mother and child count. Geneva, World Health Organization, 2005. (http://www.who.int/whr/2005/en/index.html) ## 214 cThe WHO Global Roll Back Malaria database. (http://www.who.int/globalatlas/autologin/malaria_login.asp) ## 215 dWHO report 2005. Global Tuberculosis Control; Surveillance, Planning, Financing. Geneva, World Health Organization, 2005.(http://www.who.int/tb/publications/global_report/2005/pdf/Full.pdf) ## 216 eThe WHO Global Database on Child Growth and Malnutrition. (http://www.who.int/nutgrowthdb)\n ## ...4 ...5 Immunization coverage (%) among 1-year-oldsa ...7 ...8 Antenatal ## 211 <NA> NA <NA> <NA> <NA> <NA> ## 212 <NA> NA <NA> <NA> <NA> <NA> ## 213 <NA> NA <NA> <NA> <NA> <NA> ## 214 <NA> NA <NA> <NA> <NA> <NA> ## 215 <NA> NA <NA> <NA> <NA> <NA> ## 216 <NA> NA <NA> <NA> <NA> <NA> ## ...10 Births attended by ...12 Contraceptive ...14 Children under-5 using ## 211 <NA> <NA> <NA> <NA> <NA> <NA> ## 212 <NA> <NA> <NA> <NA> <NA> <NA> ## 213 <NA> <NA> <NA> <NA> <NA> <NA> ## 214 <NA> <NA> <NA> <NA> <NA> <NA> ## 215 <NA> <NA> <NA> <NA> <NA> <NA> ## 216 <NA> <NA> <NA> <NA> <NA> <NA> ## ...16 TB detection ...18 TB treatment ...20 Antiretroviral ## 211 <NA> <NA> <NA> <NA> <NA> <NA> ## 212 <NA> <NA> <NA> <NA> <NA> <NA> ## 213 <NA> <NA> <NA> <NA> <NA> <NA> ## 214 <NA> <NA> <NA> <NA> <NA> <NA> ## 215 <NA> <NA> <NA> <NA> <NA> <NA> ## 216 <NA> <NA> <NA> <NA> <NA> <NA>
colnames(dat)
## [1] "...1" ## [2] "...2" ## [3] "...3" ## [4] "...4" ## [5] "...5" ## [6] "Immunization coverage (%) among 1-year-oldsa" ## [7] "...7" ## [8] "...8" ## [9] "Antenatal" ## [10] "...10" ## [11] "Births attended by" ## [12] "...12" ## [13] "Contraceptive" ## [14] "...14" ## [15] "Children under-5 using" ## [16] "...16" ## [17] "TB detection" ## [18] "...18" ## [19] "TB treatment" ## [20] "...20" ## [21] "Antiretroviral"
str(dat)
## 'data.frame': 216 obs. of 21 variables: ## $ ...1 : num NA NA NA NA NA 1 2 3 4 5 ... ## $ ...2 : logi NA NA NA NA NA NA ... ## $ ...3 : chr NA "Country" NA NA ... ## $ ...4 : chr "WHO" "region" NA NA ... ## $ ...5 : logi NA NA NA NA NA NA ... ## $ Immunization coverage (%) among 1-year-oldsa: chr "Measles" NA NA "2003" ... ## $ ...7 : chr "DTP3" NA NA "2003" ... ## $ ...8 : chr "HepB3" NA NA "2003" ... ## $ Antenatal : chr "care coverageb" "(%)" NA NA ... ## $ ...10 : chr NA "year" NA NA ... ## $ Births attended by : chr "skilled health personnelb" "(%)" NA NA ... ## $ ...12 : chr NA "year" NA NA ... ## $ Contraceptive : chr "prevalence rateb" "(%)" NA NA ... ## $ ...14 : chr NA "year" NA NA ... ## $ Children under-5 using : chr "insecticide-treated netsc" "(%)" NA NA ... ## $ ...16 : chr NA "year" NA NA ... ## $ TB detection : chr "rate under DOTSd" "(%)" NA NA ... ## $ ...18 : chr NA "year" NA NA ... ## $ TB treatment : chr "success under DOTSd" "(%)" NA NA ... ## $ ...20 : chr NA "year" NA NA ... ## $ Antiretroviral : chr "therapy coveragee" "(%)" NA "Dec 2004" ...
Subsetting
$operator allows us to pick out a column by name- A new object can be created from the column
- The column can also be modified with this operator
- Concept of getting and setting
# Get anc <- dat$Antenatal
To see value of object bound to anc, run the name alone
anc
## [1] "care coverageb" "(%)" NA NA ## [5] NA "52" "81" "79" ## [9] "..." "..." "..." "..." ## [13] "82" "..." "..." "70" ## [17] "..." "63" "39" "89" ## [21] "..." "..." "..." "88" ## [25] "..." "84" "99" "99" ## [29] "84" "..." "..." "72" ## [33] "93" "44" "77" "..." ## [37] "..." "…" "51" "..." ## [41] "..." "90" "87" "..." ## [45] "..." "..." "84" "..." ## [49] "..." "..." "…" "98" ## [53] "72" "..." "..." "..." ## [57] "100" "56" "54" "..." ## [61] "..." "..." "..." "27" ## [65] "..." "..." "..." "94" ## [69] "92" "91" "..." "90" ## [73] "..." "..." "86" "74" ## [77] "89" "88" "79" "..." ## [81] "..." "..." "65" "97" ## [85] "..." "..." "..." "..." ## [89] "..." "..." "..." "99" ## [93] "82" "88" "..." "83" ## [97] "88" "44" "..." "..." ## [101] "91" "..." "..." "..." ## [105] "..." "91" "94" "..." ## [109] "98" "53" "..." "..." ## [113] "63" "..." "..." "..." ## [117] "..." "..." "…" "71" ## [121] "..." "85" "..." "49" ## [125] "..." "..." "85" "39" ## [129] "61" "..." "..." "77" ## [133] "36" "..." "..." "..." ## [137] "..." "85" "94" "..." ## [141] "..." "62" "..." "99" ## [145] "89" "96" "93" "..." ## [149] "..." "..." "..." "..." ## [153] "91" "77" "82" "..." ## [157] "..." "82" "..." "..." ## [161] "..." "..." "..." "89" ## [165] "..." "..." "..." "91" ## [169] "..." "..." "..." "..." ## [173] "75" "..." "..." "..." ## [177] "78" "..." "96" "..." ## [181] "67" "87" "..." "92" ## [185] "90" "97" "..." "96" ## [189] "..." "..." "95" "..." ## [193] "..." "70" "34" "94" ## [197] "82" NA NA NA ## [201] NA "70" "84" "66" ## [205] "84" "46" "77" NA ## [209] NA NA NA NA ## [213] NA NA NA NA
Hmmm, it’s rather long. Let’s shorten the output a little.
head(anc)
## [1] "care coverageb" "(%)" NA NA ## [5] NA "52"
tail(anc)
## [1] NA NA NA NA NA NA
- These objects we have extracted from our data frame are known as atomic vectors.
[]and[[]]also allow sub-setting with numerical indices- Getting/setting applies
anc2 <- dat[['Antenatal']] anc3 <- dat[[9]]
identical(anc, anc2) identical(anc, anc3)
## [1] TRUE ## [1] TRUE
Data Cleaning
- Time-consuming
- Development of elegant constructs
- R scripts allow documentation of cleaning data steps
- Scripting: quality assurance, knowledge sharing and correction/reversal where necessary
- Scripting -> package development -> Automation
- 3rd party packages available (limited value)
Attempt to improve data import (DEMO)
Simple summary statistics
summary()- For categorical data -
table() - For numerical data -
fivenum()
Visualization
- Outstanding graphics capabilities
- Base R, other packages like lattice and ggplot2
- Run
demo("graphics")anddemo("image")
Concluding your Session
- You will be prompted to save your session.
- 99% of the time DON’T do this!
- Rather than saving objects, save the scripts that created them.
- You can save individual objects with the function
saveRDS() - To recall the object, used
readRDS().
# Create a variable with the filename. Why? myfile <- "who_data.rds" # Save the object on disk saveRDS(dat, file = myfile) # Read the object from disk diskdat <- readRDS(file = myfile) # Check if it's the same as the earlier object identical(dat, diskdat)
## [1] TRUE
Resources
- Quick R - https://www.statmethods.net/
TO DO
- Pick any old Excel file and import it into R. Give it the name
mydat:- You may have to install the readxl package.
- What is the result of
nrow(mydat)? - what is the result of
ncol(mydat)? - Run
lapply(mydat, typeof)and inspect the result. What do you think happened? (Tip: You are free to check?typeofand?lapply, or even Google!)