#My notes on R

• help.search(‘mean’) OR ?mean — to get help on mean
• Data types: int, decimal, number, vector et al

• A vector of like objects
• is.atomic
• is.vector == TRUE
• is.list == FALSE

##List:

• A bag of unlike objects

• NOT is.atomic

• is.vector == TRUE

• is.list == TRUE

• All numeric datatypes are Number unless specified with nL (100L) for integer

• Creating a list/vector “x<-c(1, ‘abc', 55);”, element of 3 with different elements in the list/vector; Do NOT use double quotes

• Creating vector is “x <- vector("numeric", length = 10)"

• Create a list with a sequence “x<-1:20”'

• Casting is done by doing as.numeric(x) or as.logical(x) or as.character(x)

• '''m<-matrix(1:6, nrow=2, ncol=3); dim (m) OR attributes(m) — shows dimensions'''

  • matrix with column & row binding — cbind and rbind
  • x<-1:3; y<-10:12; cbind(x,y)
  • rbind(x,y)

• x <- list(1,”a”, TRUE, 1 + 4i) — creates a list of objects (more like a bag)

• Factors are used for representing categorical data. x<- factor(c('yes','no','yes','no','no'),levels=c('yes','no'))

  • To obtain numerical data do unclass(x)

• is.na() vs is.nan()

• data frames are tables for data; Elements of the column have the same length;

  • data frames have row.names
  • Data Frames are usually created by calling read.table() or read.csv()
  • Can be converted to a matrix by calling data.matrix()
  • x<-data.frame(foo=1:4, bar=c(T,T,F,F))
  • nrow(x) = 4 and ncol(x) = 2

• Vector can have names x<-1:2; names(x)<-c(‘one’,’two’,’three’);

• List can have names

• matrices can have names as well called dimnames. dimnames(m)<-list(c(‘a’,’b’), c(‘c’,’d’))

• Find out function arguments str(split)

• [ is for extracting from an object
  • x[1]; x[1:4]; x[x > ’a’]; u<- x > ‘a’, x[u] — get all the elements that have TRUE in u
• [[ is used for a matrix
  • x[1,2], x[,1]
• $ is extracting with name
• Print first two rows x[1:2,]
• x$Ozone[is.na(x$Ozone)] — selecting column that is NA
• mean(x$Ozone[!is.na(x$Ozone)]) — mean exclude NA
• y<-subset(x,x$Ozone>31 & x$Temp>90) — Create a subset of a table with given conditions
• max(y$Ozone[!is.na(y$Ozone)]) — max of ozone

• read.table, read.csv

• readLines

• source — reading R code files (inverse of dump)

• dget for getting R code files (inverse of dput)

• load for reading in saved workspaces

• unsearialize, for reading R objects in binary form

• write.table, writeLines, dump, dput, save, serialize

• read.table

  • file, the name of a file, or a connection
  • header, logical indicating if the file has a header line
  • sep, a string indicating how the columns are separated
  • colClasses, a character vector indicating the class of each column in the dataset · nrows, the number of rows in the dataset
  • comment.char, a character string indicating the comment character
  • skip, the number of lines to skip from the beginning
  • stringsAsFactors, should character variables be coded as factors?

• Efficient read.table

  • initial <- read.table("datatable.txt", nrows = 100) classes <- sapply(initial, class)
  • tabAll <- read.table("datatable.txt",colClasses = classes)

• dput(trees) or dput(trees, files=‘trees.R’) # efficient way for recreating

• dump is dput for multiple objects

  • dump(c(‘x’,’y’), file=‘x_y.R’)

• file opens a file

• gzfile opens a compressed gzip

• bzfile

• url opens a connection to a webpage

• con<-file(‘abc.txt’, ‘r’)

• data<-read.csv(con)

• close(con)

• simpler read.csv(‘abc.csv’,’r’)

for loop: for(i in 1:10) { … } if: y<-if(…) { 55 } else { 93 } vector is an associative array. Index can be number or character. seq_len(5) —> creates a sequence next — similar to continue

##Functions:

formals(function_name) # describes arguments of a function
args(function_name)  # prints all the arguments;  a better method
# Return value is the last expression in the function.
f<-function(a,b=1,c=TRUE){ … } # function name is ‘f’
ls(environment(fun1)) # This will tell the variables used inside function fun1

#Date:

as.Date
x<-Sys.time()
p<-as.POSIXlt(x)
unclass(p)

#apply, lapply, sapply

• lapply: Loop over a list and evaluate a function on each element · sapply: Same as lapply but try to simplify the result
• apply: Apply a function over the margins of an array
• tapply: Apply a function over subsets of a vector
• mapply: Multivariate version of lapply
• apply is used for doing the same operation to the margins of the matrix
  • m<-matrix(c(1:9),nrow=3,ncol=3) # create a 3x3 matrix
  • Apply takes three parameters,
    • M — matrix
    • margin — 1 column wise & 2 row wise
    • function that needs to be applied

m<-matrix(c(1:9),nrow=3,ncol=3) # create a 3x3 matrix
apply(m,1,mean)  # mean column wise
apply(m,2,function(x) x^2)  # square all the elements, row wise; 

• lapply & sapply is used for doing the same operation on every element of a list

l <- c(1:9) # create a vector of 1 to 9
lapply(l, function(x) x^2) # squaring all the elements of the vector
# sapply behaves the same way as lapply but prints better
x <- list(a = 1:4, b = rnorm(10), c = rnorm(20, 1), d = rnorm(100, 5))
lapply(x,mean) # prints the mean of a, b, c, d
sapply(x, mean) # prints it well
x <- list(a = matrix(1:4, 2, 2), b = matrix(1:6, 3, 2))
lapply(x,function(e) e[,1]) # Anonymous function for extracting the first column

# tapply is used to apply a function over subsets of a vector
x <- c(rnorm(10), runif(10), rnorm(10, 1))
f<-gl(3,10)   # create 3 groups of 10, Level 1, 2 & 3
tapply(x, f, mean)  # calculate mean for each group
f2<-int.rep(1:3,10) # creates 3 groups of 1,2,3
tapply(x,f2,mean)   # calculate mean for each group

• split takes a vector or other objects and splits it into groups determined by a factor or list of factors

  • split(x,f,drop=FALSE,…)
  • x is a Vector
  • f is factor or list of factors
  • drop indicates whether empty factors levels should be dropped

x <- c(rnorm(10), runif(10), rnorm(10, 1))
f<-gl(3,10) # creates a vector with 10 of 1, 10 of 2 and 10 of 3
split(x,f)  # splits the list x into 3 parts with first 10 values in one, second 10 in another and so on
lapply(split(x,f), mean) # calculate means on those

library(datasets)
head(air quality)
# Group values by month
s<-split(airquality,airquality$Month)

# print the value mean of Ozone, Solar.R and Wind
sapply(s, function(x) colMeans(x[, c("Ozone", "Solar.R", "Wind")]))
# print the value mean of Ozone, Solar.R and Wind; Exclude NA values
sapply(s, function(x) colMeans(x[, c("Ozone", "Solar.R", "Wind")], na.rm = TRUE))

• To have more than one variable in group by

# For grouping by multiple columns you can use list
# eg: If there is a factor Male/Female & another factor Month
x<-rnorm(10)   # simulating readings
gender<-gl(2,5)
month<-gl(5,2)

str(split(x,list(gender,month)))   # has empty rows

# drop empty rows
str(split(x,list(gender,month),drop=TRUE))
# in the printed value, 1.1 means Male & Month ()

• mapply is a multivariate apply of sorts which applies to a function in parallel over a set of arguments

list(rep(1, 4), rep(2, 3), rep(3, 2), rep(4, 1))
# functionally same
* mapply(rep, 1:4, 4:1)

#Debugging

• invisible(x) # do not print the value of x even if the function returns x
• traceback: prints out the function call stack after an error occurs; does nothing if there’s no error
• debug: flags a function for “debug” mode which allows you to step through execution of a function one line at a time
• browser: suspends the execution of a function wherever it is called and puts the function in debug mode
• trace: allows you to insert debugging code into a function a specific places
• recover: allows you to modify the error behavior so that you can browse the function call stack
• debug(lm) # stop when method lm is called
• lm(x - y) # this will stop in lm and show the code in the browser

#STR

• str function on any data gives very informative output
• summary function is another that outputs statistically

rnorm(n=number of observations,mean,sd)

• The numbers are pseudo random. The numbers can be recreated by setting the seed.

# This will yield the same 5 ranom numbers every time
set.seed(1)
rrand(5,10)

• poisson's distribution

rpois(n=number of random values,lambda=vector of (non-negative) means)

• random sampling

sample(1:100,5)
sample(letters,5)

##Profiling

• system.time is quite handy

system.time(function);
system.time({
...
});

https://github.com/edj-boston/coursera-r-programming/blob/master/week-02/quiz.md