R Programming-Short Notes

# R Standard Formulas

getwd()

setwd()

install.packages("package_name")

library("package_name")

require("package_name")

help(attribute)

?attributes

#Reading Different Files

data <- read.csv(file="file_name.csv",header = True,stringsAsFactors = False)

#Exploring the Data

ncol(data)

nrow(data)

str(data)

summary(data)

View(data)

colnames(data)

#Sub setting Data

data <- data[,-c(1,2,4:7)] # Eliminating the columns

#checking missing values

sapply(data,function(x) sum(is.na(x)))

#removing the rows with missing value

data <- na.omit(data)

#Converting variables

data$attribute < - as.numeric(data$arribute1)

# Continuous variable exploration

plot(data$variable)

hist(data$variable)

#Visualizing variable by applying log

hist(log10(data$variable))

#Categorical Variables Exploration

table(data$variable)

table(data$variable)

#Bivariate Categorical Variable Analysis

table(data$variab,data$Churn)

table(data$MultipleLines,data$Churn)

#Bivariate Continous Variable Analysis

BiVarDataCont = group_by(data, Churn)

summarise(BiVarDataCont, avg_value = mean(tenure))

summarise(BiVarDataCont, avg_value = mean(MonthlyCharges))

aggregate(tenure~Churn,data,mean)

# Creating buckets

data$attribute_band <- ifelse(data$attribute>750,"750+",

ifelse(data$attribute >720,"720-750","<720"))

#Final variable selection in model

data <- data[,-c(5,7,8,9,10:15)]

#Co-relation matrix

cor_mat <- cor(data)

#In general,after analyzing correlated variables (Positive and negative correlation) variable that have correlation index more than 0.7 are removed