CS 215 Session : Titanic Data Analysis Script

titanic.sce

//0. Read the dataset!

//1. loading dataset
data=read_csv("~/cs215/session_22_07/train.csv");
header = data(1, :); //save the header for description
data=evstr(data(2:$, :));
ageIn = 4;
genderIn = 3;
classIn = 2;
fareIn = 7;
survivorIn = 9;

//2. Understanding the dataset

//Age distribution
scf(1);
histplot(10, data(:, ageIn), style=4);
title("Distribution Of Age");
xlabel("Age");
//Gender distribution
scf(2);
pie(histc(2, data(:, 3)), ["female", "male"]);
title("Distribution Of Gender");

//class distribution

//Profile of an average passenger, median passenger?

average_passenger=[header;string(mean(data, 'r'))]; //is this even useful?


//3. Asking simple questions
//What % of people survived?
perc_survived = (size(find(data(:, survivorIn) == 1)', 1) / size(data, 1)) * 100;
//Average age of survivors?
avg_survivor=mean(data((find(data(:, survivorIn) == 1)'), 4), 'r');
//Dominant surviving gender?

//4. Lets find out which factors were the most important

std_age = stdev(data(:, ageIn));
std_gender = stdev(data(:, genderIn));
std_class = stdev(data(:, classIn));
std_fare = stdev(data(:, fareIn));
std_survivor = stdev(data(:, survivorIn));


//corr coefficient(x, y) = sigma(x, y) / sigma(x) * sigma(y)
sigma_age_sur=corr(data(:, ageIn), data(:, survivorIn), 1);
corr_age_sur = sigma_age_sur / (std_age * std_survivor)

sigma_gender_sur=corr(data(:, genderIn), data(:, survivorIn), 1);
corr_gender_sur = sigma_gender_sur / (std_gender * std_survivor)

sigma_class_sur=corr(data(:, classIn), data(:, survivorIn), 1);
corr_class_sur = sigma_class_sur / (std_class * std_survivor)

sigma_fare_sur=corr(data(:, fareIn), data(:, survivorIn), 1);
corr_fare_sur = sigma_fare_sur / (std_fare * std_survivor)

train.csv