Titanic-Survival Test
Aafiya & Neha
4/1/2021
Introduction
The sinking of the RMS Titanic is one of the most infamous shipwrecks in history. On April 15, 1912, during her maiden voyage from Southampton to New York, the Titanic sank after colliding with an iceberg, killing 1502 out of 2224 passengers and crew. This sensational tragedy shocked the international community and led to better safety regulations for ships.
One of the reasons that the shipwreck led to such loss of life was that there were not enough lifeboats for the passengers and crew. Although there was some element of luck involved in surviving the sinking, some groups of people were more likely to survive than others, such as women, children, and the upper-class.
This mini-project involves the analysis of what sorts of people were likely to survive.
Creating an Environment
- This stage involves :-
- Storing training dataset in titanic.train and testing dataset in titanic.test
- Loading other required libraries to this environment.
setwd("C:/Users/SUNIL/Desktop/CS/Semester6/Data Science/Project")
library(ggplot2)## Warning: package 'ggplot2' was built under R version 4.0.4titanic.train =read.csv("train.csv",header=TRUE)
titanic.test =read.csv("test.csv",header=TRUE)
names(titanic.train)## [1] "PassengerId" "Survived" "Pclass" "Name" "Sex"
## [6] "Age" "SibSp" "Parch" "Ticket" "Fare"
## [11] "Cabin" "Embarked"names(titanic.test) ## [1] "PassengerId" "Pclass" "Name" "Sex" "Age"
## [6] "SibSp" "Parch" "Ticket" "Fare" "Cabin"
## [11] "Embarked"Creating complete dataset
- Creating dataframe with complete data (containing all rows of train and test set).
ncol(titanic.test)==ncol(titanic.train) ## [1] FALSEtitanic.test$Survived<-NA
titanic.complete=rbind(titanic.train,titanic.test)Cleaning the data
a. Checking summary (for NA’s in our dataset)
summary(titanic.complete)## PassengerId Survived Pclass Name
## Min. : 1 Min. :0.0000 Min. :1.000 Length:1309
## 1st Qu.: 328 1st Qu.:0.0000 1st Qu.:2.000 Class :character
## Median : 655 Median :0.0000 Median :3.000 Mode :character
## Mean : 655 Mean :0.3838 Mean :2.295
## 3rd Qu.: 982 3rd Qu.:1.0000 3rd Qu.:3.000
## Max. :1309 Max. :1.0000 Max. :3.000
## NA's :418
## Sex Age SibSp Parch
## Length:1309 Min. : 0.17 Min. :0.0000 Min. :0.000
## Class :character 1st Qu.:21.00 1st Qu.:0.0000 1st Qu.:0.000
## Mode :character Median :28.00 Median :0.0000 Median :0.000
## Mean :29.88 Mean :0.4989 Mean :0.385
## 3rd Qu.:39.00 3rd Qu.:1.0000 3rd Qu.:0.000
## Max. :80.00 Max. :8.0000 Max. :9.000
## NA's :263
## Ticket Fare Cabin Embarked
## Length:1309 Min. : 0.000 Length:1309 Length:1309
## Class :character 1st Qu.: 7.896 Class :character Class :character
## Mode :character Median : 14.454 Mode :character Mode :character
## Mean : 33.295
## 3rd Qu.: 31.275
## Max. :512.329
## NA's :1Here, we observed that the Age, Survived, Fare columns have NA’s in them
1. Handling empty string
- Embarked column has empty string that need to be handled.
table(titanic.complete$Embarked)##
## C Q S
## 2 270 123 914which(titanic.complete$Embarked=="")## [1] 62 830titanic.complete[titanic.complete$Embarked=="","Embarked"]<-"S"
table(titanic.complete$Embarked)##
## C Q S
## 270 123 9162. Handling NA’s
a. Handling NA’s in Age Column
- In this phase,data having NA value in Age column will be assigned with median of ages(exculding NA’s)
- For complete set.
table(is.na(titanic.complete$Age)) ##
## FALSE TRUE
## 1046 263median_Age<-median(titanic.complete$Age,na.rm = TRUE)
titanic.complete$Age[is.na(titanic.complete$Age)]<-median_Age- For train set.
table(is.na(titanic.train$Age)) ##
## FALSE TRUE
## 714 177median_Age<-median(titanic.train$Age,na.rm = TRUE)
titanic.train$Age[is.na(titanic.train$Age)]<-median_Ageb. Handling NA’s in Survived Column
- In this phase,data having NA value in Survived column will be assigned with 0 value.
- For complete set.
table(titanic.complete$Survived)##
## 0 1
## 549 342table(is.na(titanic.complete$Survived))##
## FALSE TRUE
## 891 418titanic.complete$Survived[is.na(titanic.complete$Survived)]<-0- For test set.
table(titanic.train$Survived)##
## 0 1
## 549 342table(is.na(titanic.train$Survived))##
## FALSE
## 891titanic.train$Survived[is.na(titanic.train$Survived)]<-0c. Handling NA’s in Fare Column
- In this phase,data having Na’s in Survived column will be assigned with median of fare(exculding NA’s).
table(is.na(titanic.complete$Fare))##
## FALSE TRUE
## 1308 1median_Fare<-median(titanic.complete$Fare,na.rm = TRUE)
titanic.complete$Fare[is.na(titanic.complete$Fare)]<-median_FareData Exploration
1. Sex : Survived and Deceased Comparison
counts <- table(titanic.complete$Survived, titanic.complete$Sex)
counts##
## female male
## 0 233 734
## 1 233 109rownames(counts)<-c("Not Survived","Survived")
survival_precentage<-c(counts[2]/(counts[1]+counts[2]),counts[4]/(counts[3]+counts[4]))
survival_precentage## [1] 0.5000000 0.1293001barplot(counts,main = "Sex Survival Plot",col = c("red","green"))
legend("topleft",c("Not survived","Survived"),fill = c("red","green"))
Inference
- Among passengers in Titanic- 64% are Males and 36% are Females.
- But the analysis indicates that Females(50%) have a better survival rates than Males(13%).
2. Pclass : Survived and Deceased Comparison
Pclass_survival<-table(titanic.complete$Survived, titanic.complete$Pclass)
rownames(Pclass_survival)<-c("Not Survived","Survived")
Pclass_survival##
## 1 2 3
## Not Survived 187 190 590
## Survived 136 87 119survival_precentage<-c(Pclass_survival[2]/(Pclass_survival[1]+Pclass_survival[2]),Pclass_survival[4]/(Pclass_survival[3]+Pclass_survival[4]),Pclass_survival[6]/(Pclass_survival[5]+Pclass_survival[6]))
survival_precentage## [1] 0.4210526 0.3140794 0.1678420barplot(Pclass_survival,main = "PClass Survival Plot",xlab="Passenger Class",col = c("red","green"))
legend("topleft",c("Not survived","Survived"),fill = c("red","green"))
Inference
- Majority of the passengers travelled in 3rd Class.
- The analysis indicates that the passengers travelling in higher classes are more likely to decease.
3.Sibsp : Survived and Deceased Comparison
Sibsp_survival<- table(titanic.complete$Survived,titanic.complete$SibSp)
rownames(Sibsp_survival)<-c("Not Survived","Survived")
Sibsp_survival##
## 0 1 2 3 4 5 8
## Not Survived 681 207 29 16 19 6 9
## Survived 210 112 13 4 3 0 0barplot(Sibsp_survival,main = "Number of siblings/spouses aboard-Survival Plot",xlab="Sibling/Spouse",col = c("red","green"))
legend("topright",c("Not survived","Survived"),fill = c("red","green"))
Inference
- Majority of people with 0 siblings/spouse deceased.
- Its evident that those who came without thier siblings/spouse deceased the most.
4.Parch : Survived and Deceased Comparison
Parch_survival<- table(titanic.complete$Survived,titanic.complete$Parch)
rownames(Parch_survival)<-c("Not Survived","Survived")
Parch_survival##
## 0 1 2 3 4 5 6 9
## Not Survived 769 105 73 5 6 5 2 2
## Survived 233 65 40 3 0 1 0 0barplot(Parch_survival,main = "Number of Parents/children aboard-Survival Plot",xlab="Parents/Children",col = c("red","green"))
legend("topright",c("Not survived","Survived"),fill = c("red","green"))
Inference
- Majority of people with 0 Parent/Children deceased.
- Its evident that those who didn’t come with their parents/children were the one who deceased.
5. Embarked : Survived and Deceased Comparison
Embarked_survival<- table(titanic.complete$Survived,titanic.complete$Embarked)
rownames(Embarked_survival)<-c("Not Survived","Survived")
Embarked_survival##
## C Q S
## Not Survived 177 93 697
## Survived 93 30 219survival_precentage<-c(Embarked_survival[2]/(Embarked_survival[1]+Embarked_survival[2]),Embarked_survival[4]/(Embarked_survival[3]+Embarked_survival[4]),Embarked_survival[6]/(Embarked_survival[5]+Embarked_survival[6]))
survival_precentage## [1] 0.3444444 0.2439024 0.2390830barplot(Embarked_survival,main = "Coaches Embarked-Survival Plot",xlab="Embarkment",col = c("red","green"))
legend("topleft",c("Not survived","Survived"),fill = c("red","green"))
Inference
- Among passengers in Titanic-20.6% people embarked from Cherbourg, 11.83% people embarked from Queenstown and 69.97% people embarked from Southampton.
- But the analysis indicates that people embarked from Cherbourg(34.44%) have a better survival rates than who embarked from Queenstown(24.39%) and Southampton(29.90%).
6. Age Survival(Range-wise) : Survived and Deceased Comparison
Ages<-table(titanic.complete$Survived,cut(titanic.complete$Age, breaks=c(0,14,24,64,Inf)))
rownames(Ages)<-c("Not Survived","Survived")
colnames(Ages)<-c("Children","Youth","Adults","Seniors")
Ages##
## Children Youth Adults Seniors
## Not Survived 62 228 665 12
## Survived 45 73 223 1barplot(Ages,main = "Ages Survival(Range-wise) Plot",xlab="Age Category",col = c("red","green"))
legend("topright",c("Not survived","Survived"),fill = c("red","green"))
Inference
- Majority of the passengers are adults.
- Majority of the survivors are children.
- It is evident that passengers in senior category has least survived.
7. Ticket-Fare Survival(Range-wise) : Survived and Deceased Comparison
fcount<-table(titanic.complete$Survived,cut(titanic.complete$Fare, breaks=c(-Inf,10,20,30,Inf)))
rownames(fcount)<-c("Not Survived","Survived")
colnames(fcount)<-c("<$10","$10-$20","$20-$30",">$30")
fcount##
## <$10 $10-$20 $20-$30 >$30
## Not Survived 424 186 150 207
## Survived 67 76 63 136barplot(fcount,main = "Ticket-Fare Survival Plot",xlab="Ticket-Fare Range",col = c("red","green"))
legend("topright",c("Not survived","Survived"),fill = c("red","green"))
Inference
- Majority of the passengers have purchased the ticket for a fare less than $10.
- Majority of the survivors have purchased the ticket for a fare greater than $30.
- It is evident that passengers with an expensive ticket are more likely to survive when compared with the passengers with lower ticket fare.
8. Age-Sex Survival: Survived and Deceased Comparison
ggplot(titanic.complete, aes(x=Age, fill=factor(Survived))) +geom_histogram(bins=30)+ facet_grid(.~Sex)+
ggtitle("Age-Sex Survival")+
scale_fill_discrete(name="Survived")
Inference
- Majority of the deceased passengers were Male.
- Majority of the survivors were Female.
- It is evident that male passengers in age group of 20-35 yrs have deceased the most and women in age group 20-30 have survived the most.
9. Fare-Sex Survival: Survived and Deceased Comparison
ggplot(titanic.complete, aes(x=Fare, fill=factor(Survived))) +geom_histogram(bins=30)+ facet_grid(.~Sex)+ggtitle("Fare-Sex Survival")+ scale_fill_discrete(name="Survived")
Inference
- Majority of the deceased passengers were Male.
- Majority of the survivors were Female.
- It is evident that male passengers who had a ticket in between $0-$25 have deceased the most and women who had a ticket in between $0-$25 have survived the most.
Analysis
- Most of the Females survived as compared to Males.
- Most of the passengers with 1st class survived.
- Most of the passengers who embarked from Southampton deceased.
- Most of the adults have deceased.
- Most of the passengers having more than $30 ticket survived while ones having less than $10 deceased.
- Males in range of 20-35 yrs have deceased the most and female under 20-30 yrs have survived the most.
Hypothesis Testing
by(titanic.train$Age,titanic.train$Survived,mean)## titanic.train$Survived: 0
## [1] 30.02823
## ------------------------------------------------------------
## titanic.train$Survived: 1
## [1] 28.29143Ho=The Titanic survivors were younger than the passengers who died.
t.test(Age~Survived,var.equal=TRUE,data=titanic.train)##
## Two Sample t-test
##
## data: Age by Survived
## t = 1.9395, df = 889, p-value = 0.05276
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
## -0.02074894 3.49434975
## sample estimates:
## mean in group 0 mean in group 1
## 30.02823 28.29143