A Pelican Blog

Titanic Dataset Visualization

Link to Visualization

Summary

This visualization describes the percentage of passengers that surived during the Titanic disaster. The circles represent the susbet of passengers categorized by sex, passenger class and point of embarkation. The size of the circle corresponds to the percent value; the higher the proportion of survivers, the larger the circle.

Design

This purpose of this visualization is to elegantly convey relationships between the survival rates amongst different categories of passengers. This message is delivered with the size of each circle. The circles are lined up horizontally since the natural tendancy for most people is to glance at a visual from left to right. There is enough space in between each circle so that they don't overlap but close enough that they can all fit on a regular computer monitor screen. The radius of the circle is aligned with the square root of percentage of survivors so that each circle, then multiplied by a factor of 8 to make them large enough to be visible. This is a reasonable choice since the cirlces are large enough to be visible yet small enough for them all to fit on the a single screen. The colors of the circle and the text are paste blue and white, respectively. This combination was selected because it does not strain the eyes when looking at it for a long time. The font of the text is the standard sans-serif and the size of the font is small enough to fit within all of the circles, yet large enough to be readable. When hovering over each cirlce, the tooltip feature is used to display the labels. The labels states the category of the passengers and the associated survival rate, as well as the actual count of survivors. The color of box is light pink, which was selected due to its contrast with the blue circles. Both are easy on the eyes separately and when displayed together. Because of the slight opacity of the labels, font color was chosen to be black so that it can be clearly read. In order to create a visual separation between the categories, a vertical line is added between the sex, passenger class and port of embarkation categories.

Feedback

Feedback 1

Noticed that gender, passenger class and port of embarkation were being compared
Noticed that survival rate was higher in females than males and was correlated with passenger class
Suggestion was to use pastel palette instead of dark colors since they are too bright.
o Circle color changed
Questioned as to the meaning of the port of embarkation letters
o Addressed in the Data Story section

Feedback 2

Noticed circles of different sizes lined across a horizontal line with labels
Noticed that hover certain circles are larger than others
Upon looking at the details while hovering over the circles, noticed that the size of the circle was related to the survival rate
Suggestion was to add a title so that it can be made more clear what we are looking at
o Title added to visualization
Noticed that females had a higher survival rate than males, as well as a decreasing survival rate when moving from passenger class 1 to 2 to 3
Question was raised as to why Port C had a larger survival rate than the other ports
o Addressed in the Data Story section

Feedback 3

Noticed that there are different categories for survival rates: gender, passenger class and port of embarkation
Noticed that the size of the circle is used to visually convey the survival rates relative to each other
Noticed that females had the largest survival rate; even beating out those in passenger class one
A question was raised as to the meaning of point of embarkation
o Since this issue came up 2nd time, text was added below the visualization that explained what the ports letters meant (S, C and Q → Southampton, Cherbourg, Queenstown)

Feedback 4

Include the actual count of passengers in the tooltip under the percentage of survivors
o Count of survivors added
Instead of using the actual percentage values for the radius, use the square root of the values
o Changed
Separate the circles for the different categories
o Vertical lines added between the categories to create separation

Titanic Dataset – Story

In this visualization, the survival rate of the passengers was examined for the following categories:
• Sex
• Passenger Class
• Point of Embarkation
As we all know, there were a limited number of lifeboats that could only accommodate some of the passengers. These passengers that were boarded onto the lifeboats were selected by the crewmembers and there was a clear bias towards who got to board and who got left behind. The most obvious difference is the survival rate amongst males and females. Clearly, females were given priority in the lifeboat boarding process with a 74.20% survival rate amongst females, compared to a mere 18.89% survival rate amongst males. Another clear trend that we see in the survival rate is with the passenger class. The lower your class, the less chance you had at getting on a lifeboat and surviving the disaster. First class passengers had a survival rate of 62.96%, second class passengers had a survival rate of 47.28% and third class passengers had a survival rate of only 24.24%. This again makes sense since the first class passengers were more likely to be wealthy and treated better, thus the crewmembers giving the a higher priority at getting on the lifeboats. In terms of the Port of Embarkation, Ports S (Southampton), C (Cherbourg) and Q (Queenstown) had survival rates of 33.70%, 55.36% and 38.96%, respectively. Although Port C had a much higher survival rate and Ports S and Q had very similar survival rates, there is no clear indication as to why this may have happened. Some possibilities are that a larger number of first class passengers may have boarded from Port C, or could have boarded more females, or more families, or any other number of factors. This relationship can be explored with further analysis.

Financial Contributions to the 2016 Presidential Campaigns from NY

========================================================

#Load libraries
library(ggplot2)
library(lubridate)
library(dplyr)
library(tidyr)
library(gridExtra)
library(psych)

# Load the Data
data = read.csv('NYContributionData.csv', row.names = NULL)
#Convert dates from integers to DD-MM-YYYY format
data$contb_receipt_dt = dmy(data$contb_receipt_dt)

Univariate Plots Section

# Summarize the data set
dim(data)

## [1] 186976     18

names …

Data Wrangling with MongoDB – OpenStreetMap Data Case Study

Published: Wed 04 May 2016
In Data Analysis.

Summary

Map Area: Waterloo, Ontario, Canada
Problems Encountered in the Map:
After initially downloading the Waterloo area and running it against the script below, I noticed the following problems with the data:
• Inconsistent methods of describing street direction (e.g. “N” or “North”)
• Inconsistent street types (e.g. St. instead …
read more

Udacity Engagement Analysis

Loading Data from CSVs

# Import csv library
import unicodecsv

# Define function to read and store data
def read_csv(filename):
    with open(filename, 'rb') as f:
        reader = unicodecsv.DictReader(f)
        return list(reader)

enrollments = read_csv('data/enrollments.csv')
daily_engagement = read_csv('data/daily_engagement.csv')
project_submissions = read_csv('data/project_submissions.csv')

# Print out first …

Introduction to Research Methods

Published: Sat 19 September 2015
In Data Analysis.

Introduction To Research Methods

Chopsticks

A few researchers set out to determine the optimal length of chopsticks for children and adults. They came up with a measure of how effective a pair of chopsticks performed, called the "Food Pinching Performance." The "Food Pinching Performance" was determined by counting the number …
read more
Baseball Salaries Analysis
Published: Wed 25 March 2015
In Data Analysis.

The Baseball Data was used to answer the following questions:
Are salaries higher in 2015 than in 1985?
Does the mean salary increase with the year?
Do a player salaries have a higher correlation with total runs or homeruns?
```
# Import packages
import csv
import numpy as np
import pandas as …
```
read more

Other articles

Univariate Plots Section

Summary

Loading Data from CSVs

Introduction To Research Methods

Chopsticks

social