increasingly popular statistical programming language r for data science. r is popular for statistical computing and statistical analysis.

what is r ?

R is open source language and relatively easy to learn. R is handle huge data sets with incredible data visualization and graphics capabilities. In r data analysis better and faster and allowed them to make better visualizations.

Basics Of R Programming

How to install R Studio ?

• Step 1 : Download R Studio Given Link Or Already You Have Go To Next Step
• Step 2: Open .exe File In Windows OS Or Mac Click On .dmg File
• Step 3: Set Install Location
• Step 4: Select Start Menu Folder And Shortcuts Create Checklist
• Step 5: Waiting Installing process then
• Step 6: Click On Finish

How to install R Packages ?

Write Blow Code in R Studio Code Editor Or R Console

install.packages("package_name")

R Basics

Variable

#Sequences
1:50   # print numbers 1 to 50 to the console
50:1   # print numbers 50 to 1 in reverse order to console
#Sequences continued...
#seq(from = 1, to = 1, by = ((to - from)/(length.out - 1)),length.out = NULL, along.with = NULL, ...)
seq(7, 49, 7) 
 #more parameter use ?funciton_name Example: ?seq
y <- c(3, 5, 8, 1, 2)
#Examining Variables
class(y) # Return y argument character, numeric or logical class 
# Factors Variables
length(y) # Return y argument length
levels(y) #Return the value of the levels of y argument 
#table function
x <- c("a", "a", "b", "b", "b", "c")
table(x)

Programming Basics

While Loop

# While Loop
i <- 1
while (i < 6) {
print(i)
i = i+1
}

if…else statement

#if…else statement
x <- -14
if(x > 0){
print("positive x")
} else {
print("not positive x")
}

For Loop

# For Loop
x<-1:100
for(i in 1:100)
  {
    if(i%%2!=0)
      {
    print(x[i])
      }
  }

Functions

y<-function(k)
  {
    s<-sum(k)
    MEAN<-s/length(k)
    print(MEAN)
  }

Vectors

#Numeric Vectors
y <- c(20, 20, 60)
#Character vectors
datascientist <- c("Programming language","statistics","domain knowledge")
#Connecting Numeric and Character Vectors
names(y) <- datascientist
y
#Subsetting vectors
y[1:2]

Basic Data Wrangling

#data frame
df<-data.frame(x=1:2,y=3:4)
is.na(df) # Return True Or Flash
sum(is.na(df)) # Return Sum Of Null Values
#dplyr
library(dplyr) #  Data Manipulation library
data("ToothGrowth") #load built in data set
View(ToothGrowth) #view data set
?ToothGrowth

Basic Plots

bar plot

#bar plot
barplot(iris$Sepal.Length)

Scatterplots

#Scatterplots
library(ggplot2)
ggplot(iris, aes(x = Sepal.Length, y = iris$Species)) + geom_point()

Histograms

#Histograms
hist(iris$Sepal.Length)

Boxplots

#Boxplots
boxplot(iris$Sepal.Length, horizontal = T)

pie chart
#pie chart
pie(table(iris$Species))

Data import

Read CVS Data File

read.csv(file="rainfall.csv", header = TRUE, sep=",")

Read Or Load Data S, dBase, SPSS, SAS, Stata, Weka, Systat

install.packages("foreign")
library(foreign)   # used to load statistical software packages
indata <- read.spss("election2020.sav")
indataframe <- as.data.frame(indata)
str(indataframe)
summary(indataframe)
install.packages("sas7bdat")  # or install Hmisc
library(sas7bdat)
data(sas7bdat.sources)

Load Or Read ODBC Database

install.packages("RODBC")
library(RODBC)
odbcDataSources()
getwd()

Read SQL Data File

myconn <-  odbcConnectAccess("(Microsoft Access Driver(*.mdb,*.accdb)); Dbq=A.mdb")
Test <- sqlFetch(myconn, "Location")
Test

Data Read Or Load Form Site

install.packages("RCurl")
library(RCurl)
data2 <- getURL("site_url")

Exploratory Data Analysis EDA in R

#load css file from local system
mba <- read.csv("mba.csv")
#Measures of Central Tendency
mean(mba$gmat)
median(mba$gmat)
#mode
getmode <- function(x)
  {
  uniquv <- unique(x)
  uniquv[which.max(tabulate(match(x,uniquv)))]
  }
getmode(mba$gmat)
#Measures of Dispersion
var(mba$gmat)
sd(mba$gmat)
range(mba$gmat)
rangevalue <- function(x)
  {
  max(x)-min(x)
  }
rangevalue(mba$gmat)
#Measures of skewness
install.packages("moments")
library(moments)
#Measures of skewness
skewness(mba$gmat)
#Measures of Kurtosis 
kurtosis(mba$gmat)
#Graphical Representation
#Boxplot
boxplot(mba$gmat,horizontal = TRUE)
#Histogram
hist(mba$gmat)
#Barplot
barplot(mba$gmat)
str(mba)
mba$datasrno <- as.factor(mba$datasrno)
#install.packages(psych)
library(psych)
describe(mba)
# to calculate Z score
qnorm(0.950)#90%
qnorm(0.975)#95%
qnorm(0.995)
#to calculate t score
qt(0.950,772)#0.90%
#qqplot
qqnorm(mba$gmat)
qqline(mba$gmat)

Model building

Linear Regression

#Simple Linear Regression
Salary_hike_Model <- lm(Salary ~ YearsExperience, data = Salary_hike)
summary(Salary_hike_Model)
plot(Salary_hike_Model)
#multiple linear regression
model3 <- lm(price ~ speed + hd + ram + screen + ads + trend + premium, data = ComputerData[-c(1441, 1701),])
summary(model3)
avPlots(model3)

Decision Trees

Decision Trees Model Use C50 library. First Install Or Update C50 packages using install.packages(“C50”) code. set predict and independent variable and build model.

install.packages("C50")
library(C50)
company_train_model <-C5.0(company_train[,-1],company_train$company_sales)
plot(company_train_model)

Random Forest

• Random Forest develops lots of decision tree based on random selection of data and random selection of variables.
• Random Forest provides the class of dependent variable based on many trees
• As the tree are based on random selection of data as well as variables, these are random tree
• many such random trees leads to a random forest
• Two major belief
  • Most of the tree can provide correct prediction of class for most part of the data
  • the tree are making mistake at different place
• That’s why random forest based on voting. for each of the observation and them decide about the class of the observation based on poll result, it is expected to be more close to the correct classification

library(randomForest)
company_train_model <- randomForest(Sales~.,data = company_train,na.action=na.roughfix,importance=TRUE)
company_train_model
plot(company_train_model)
company_train_pred <- predict(company_train_model,company_train)

Final Outcome R For Data Science

R Programming is open source Language with large community. R have 10,000+ packages. R Is Mostly Use For Visualization And Statistical Analysis. R Is Easy Learnable Better than other language.