The RMS Titanic story has living with us through
numerous stories and films until now. Titanic was known as the one of
the world’s most celebrated ship because it was the biggest ship that
has been made by human at that time. Through its popularity, many
wealthy people, high-ranking officials and celebrities were on-board to
witness the Titanic’s Maiden Voyage along with other
2,222 passengers and crew.
Departure of Titanic on April 10, 1912
On April 15, after 5 days of sailing, Titanic has sank in the North
Atlantic after crashing an iceberg. Of the 2,222 passengers and crew on
board. More than 1,500 lost their lives in the disaster. Because of this
unfortunate event, there are so many studies and speculation about the
lack of satisfactory emergency procedures when the disaster is
happening. Such as the lifeboats shortage, the violation of unwritten
law of the sea , first-class priority, and much more.
In this document, we will try to answer some speculation with the
analysis about this unfortunate event that happened in 1912.
SibSp: number of siblings / spouses aboard the
Titanic
Parch: number of parents / children aboard the
Titanic
Ticket: Ticket number
Fare: Passenger Fare
Cabin: Cabin number
Embarked: Port of Embarkation, C =
Cherbourg, Q = Queenstown,
S = Southampton
Data Cleansing
Before we go any further, we should see the condition of our data
then cleaning the data to convert a raw dataset into a useable form to
make it more relevant with our analysis later.
Missing Values
First thing that we want to do is looking for missing values. It is
important to handle missing data because any analytic results based on a
dataset with missing values could be biased. 1https://en.wikipedia.org/wiki/Missing_data
## Age Survived Pclass Name Sex SibSp Parch Ticket
## 177 0 0 0 0 0 0 0
## Fare Cabin Embarked
## 0 0 0
nrow(titanic)
## [1] 891
The only column that has missing values is Age column.
However, the proportion of the missing values is quite large: 177 rows
of 891 rows which is about 19,9% of the data. And If we do further
scrunity with our Data Structure, we can see in the
Cabin column, there are values with “” means the
Passenger’s Cabin was not specified/known during the collection of
the data. Thus we can treat the value as missing values and check
its proportion within the column.
sort(head(table(titanic$Cabin)), decreasing=T)
##
## A10 A14 A16 A19 A20
## 687 1 1 1 1 1
There are, too, large amount of missing values (687 rows of 891) in
Cabin column which is 77% of the data. Since our goal is to
avoid biased analytics, we need to drop these 2 columns because we will
not use it for our analysis.
titanic <- subset(titanic,select=-c(Age,Cabin))
Inessential Columns
What made a column to be Inessential? In this
project, we will determine it by looking at which column that’s
already been represented with another column and which
column that doesn’t have any relation with theSurvivalcolumn by just looking at it. As
of now, by our understanding, there are 6 columns with such
characteristics: Ticket, Embarked,
Fare, Name, SibSp, and
Parch.
Reason of Removal:
Ticket: Too much unique values and it doesn’t represent
anything that related to Survivability.
Embarked: It doesn’t mean much since when the disaster
happened, all passengers were already embarked from their respective
docks.
Fare: If we talk about how cheap/expensive each fare.
It should be already represented with Pclass since the
higher the Fare, the higher Pclass that you
can get.
Name: Same as Fare, we can represent each
name Title with Pclass along with
Sex.
SibSp and Parch: Since these two columns
represent how much Passenger’s Relatives that were also on-board, then
there are chances that passenger was on-board without any
relatives.
Before we drop those columns altogether, we want to simplify them
into one new column called Relatives. This column will give
you value of No if you’re alone and
Yes if you’re with relatives no matter how
much is it.
The 1st Class were roughly 16% and
39% more likely to survive than the 2nd Class
and 3rd Class respectively. But What if we want to know
the chance between the1st Classand
the2nd & 3rdaltogether?
# Change the 2nd & 3rd Class into 'Other Class'
titanic_other <- titanic
convert_class = function(x){
if(x == "1") {x <- "1st Class"}
else {x <- "Other Class"}
}
titanic_other$Pclass <- sapply(X = titanic_other$Pclass,
FUN = convert_class)
titanic_other$Pclass <- as.factor(titanic_other$Pclass)
When we grouped the other class altogether, the differences between
the 1st Class and the Other Class is about
32%. The result, however, is different with the number
that has been stated by history.com.
The reason of this might be because the dataset from Kaggle is not
complete enough which only 891 Passengers recorded out of 2222
Passengers.3https://titanicfacts.net/titanic-passengers/
However, we can clearly see the same pattern within these two
results, that, the1st Classhas
always likely to have better survival chance rather than the other
classes.
Margaret Brown, better known as ‘the unsinkable Molly Brown’, earned her
nickname because she survived the sinking of the Titanic and later went
on to become a staunch philanthropist and activist.
In this section, we am going to calculate the chance of survival
based on Passenger’s Gender to see if the
drill has been maintained when the disaster is
happening.
Women has 55% more chance to survive rather than its
counterpart so we can assume that the drill has been maintained.
However, Titanic had total of 20 Lifeboats, in which
each Lifeboat can held total of 68 people and we can conclude that the
Lifeboats could take about 1360 passengers or more.6https://en.wikipedia.org/wiki/Lifeboats_of_the_Titanic
Then we want to see how much women were on-board based on the
dataset.
#Calculate the frequency
sex_freq <- as.data.frame(table(titanic$Sex))
colnames(sex_freq)[1] <- "Sex"
knitr::kable(sex_freq)
Sex
Freq
female
314
male
577
There were only 314 Female Passengers and if we look for the actual
number, there were 402 Women Passengers on-board7https://titanicfacts.net/titanic-passengers, which were still
under the capacity of all lifeboats. Meaning that all women were
supposedly can be saved but unfortunately, they’re not.
By this, though the Birkenhead Drill was upheld as a
“chivalric ideal” among ship-crews, in fact, 60 years after the sinking
of the troopship (which was the time when RMS Titanic disaster
happened), this “ideal” was not completely adhered by the crews
on-board.
Family Factor
“Hundreds of human dramas unfolded between the order to load the
lifeboats and the ship’s final plunge: Men saw off wives and children,
families were separated in the confusion and selfless individuals gave
up their spots to remain with loved ones or allow a more vulnerable
passenger to escape.”
The Sad Parting, illustration of 1912.
This is the last factor that we want to analyze in this document. As
we know before, there were many children and women on-board along with
their Husband/Dad and there were, too, people on-board without any
relatives with them. Now we want see how much proportion of those who
were alone and those who’re with families.
There were roughly 60% people that on-board without
any relatives and 40% with their relatives. And we are
going to check their Survival Chance based on our dataset.
Passengers with relatives has a better chance of survival
(20% more) rather than people that were alone. This
might be because the number of children and wives prioritized by each
families. Moreover, if we see the “quote” from above, there were
“selfless individuals gave up their spots to remain with loved ones
or allow a more vulnerable passenger to escape” and this can be the
reason that why not all Women were saved in the first place.
Conclusion
Now we know that there are 3 factors that affect Passenger’s
Survivability:
Pclass factor : Passengers from the 1st
Class are more likely to survive in the event. This answers the
speculation that 1st Class passengers were highly
prioritized.
Sex factor : Female passenger are more
likely to survive in the event. This answers the speculation that
Female passengers were highly prioritized.
Relatives factor : Passenger that with
Relatives on-board are more likely to survive in the event.
This answers the speculation that Couples/Families were highly
prioritized.
However, there are still much speculation that we can’t answers with
our analysis about the lack of satisfactory emergency procedures and we
hope the same speculation will find some certitude in the future.
