Space Launches

Graphing for EDA (Exploratory Data Analysis)

Published

January 14, 2019

Notable topics: Graphing for EDA (Exploratory Data Analysis)

Recorded on: 2019-01-14

Timestamps by: Alex Cookson

View code

Screencast

Timestamps

str_detect

Using str_detect function to find missions with "Apollo" in their name

Starting EDA (exploratory data analysis)

fct_collapse

Using fct_collapse function to recode factors (similar to case_when function)

countrycode
countrycode

Using countrycode function from countrycode package to get full country names from country codes (e.g. "RU" becomes "Russia")

Using replace_na function to convert NA (missing) observations to "Other"

geom_line

Creating a line graph using geom_line function with different colours for different categories

fct_reorder

Using fct_reorder function to reorder factors in line graph above, in order to make legend more readable

geom_col

Creating a bar graph, using geom_col function, of most active (by number of launches) private or startup agencies

%/%

Using truncated division operator %/% to bin data into decades

complete

Using complete function to turn implicit zeros into explicit zeros (makes for a cleaner line graph)

facet_wrap

Using facet_wrap function to create small multiples of a line graph, then proceeding to tweak the graph

semi_join

Using semi_join function as a filtering step

geom_point

Using geom_point to create a timeline of launches by vehicle type

Explanation of why boxplots over time might not be a good visualization choice

geom_jitter

Using geom_jitter function to tweak the timeline graph to be more readable

Creating a second timeline graph for US vehicles and launches

Summary of screencast