Using a public database provided by NASA, this project aims to draw conclusions about spacecraft launch sites, create a model to predict the proportion of failed launches, and map the launch sites.
A. Data collection
Part 1: Get the data via GET requests and create a dataframe.
Part 2: Data cleaning/wrangling.
(rewriting the df to only contain the main columns, extracting the information from the links associted, formatting the date column, etc.)
Part 3: Filter the dataframe to only include falcon 9 releases.
Part 4: Final adjustments and file download.
(replacing missing values, downloading the file)
OUTPUT: dataset_part_1
B. Exploratory data analysis
Part 1: Calculate the number of releases on each platform.
Part 2: Calculate the number of launches for each orbit.
Part 3: Calculate the number of different landing results, distinguish successful from unsuccessful ones.
Part 4: Create a Class column using LandingOutcome.
Part 5: Using the Class column.
(calculating average, grouping by orbit and class, counting)
OUTPUT: dataset_part_2
C. Data analysis using SQL
Part 1: Upload the data to the database
Part 2: Queries with SQL
(10 exercices)
D. Data visualization
Part 1: Visualize the relationship between the flight number and the takeoff platform.
Part 2: Visualize the relationship between the mass of the load and the platform.
Part 3: Visualize the relationship between number of flights and orbit.
Part 4: Visualize the relationship between orbit and mass of the charge.
Part 5: Visualize the likelihood of success according to the platform
Part 6: Visualization of the probability of success in each month.
Part 7: Visualization of the probability of success over time.
E. Interactive Visual Analytics of launch site locations with Folium
Part 1: Mark all launching sites on a map.
Part 2: Mark on the map the successful/failed launches of each site.
Part 3: Calculate the distances between a launch site and its vicinity.