NW Assignment

My Heart Will Go On (but not yours?!?)

Today’s assignment was inspired by a student’s question, wondering if network analysis could be applied to the Titanic dataset. Instead of using the somewhat tragic dataset about surviving or perishing in the Titanic disaster, we will focus on the events leading up to it and try to solve the mystery behind why Rose didn’t help Jack from dying.

Titanic

We will use James Cameron’s 1997 film Titanic, starring Leonardo DiCaprio (as Jack) and Kate Winslet (as Rose). The film is based on a partly true story, though with a few inaccuracies.

One example is Violet Jessop, a woman who survived the RMS Olympic (1911), RMS Titanic (1912), and later the HMHS Britannic (1916), which sank after hitting a German naval mine. In the film, she was depicted as an elderly Third-Class stewardess, instead of a First-Class stewardess who was about 24 years old.

Dataset

GITHUB REPO with data and detailed instructions.

Dataset consists of two files: edges.csv & nodes.csv. The names of the file should be self-explanatory. Data was constructed from film’s script, published here, and present the cooccurrence of the characters in each scene. Both files are in the github repository of the assignment.

In edges.csv the Weight column presents the number of cooccurrences for characters, and should be used in the assignment.

Instructions:

Follow all 4 objectives in the file. For each objective, calculate and/or visualize the results, and provide explanations of WHAT you used to accomplish the task, WHY you chose that approach, and most importantly, your understanding of the results.

LET’S HYPOTHESIZE that Rose could have saved Jack, so at the end, provide your reasoning as to why she didn’t save him (there’s no right answer to this :)).

You can simply rename this file and submit it latest Friday 24:00. If you’d like, you can also create a video presentation. Making a video is a great exercise for students who are struggling with oral exams.

Over the weekend, I will review the submissions, and at the beginning of next week, I will schedule a Teams meeting where we will slowly go through the objectives together.

HINT: Use NetworkX documentation!