The Destiny 2 project that analyzes weapon stats and perks to predict their popularity
by Jorge Jimenez Garcia, as part of a Bsc in Computer Science for Technological University Dublin
-
Gather data To gather data, execute the
weaponCrawler.pyscript to generate a.d2datafile. Control the parameters of the search by modifying the 'PARAMETERS' section of the script -
Convert data to a
.csvRun thePreprocessnotebook. Note that it may need to download the database on first run or if it has been updated since the last execution. Remember to clean leftover databases (.contentfiles) -
Train models Run the
Trainnotebook. Computation may take a long time due to the use of 10-Fold Cross Validation -
Check for live performance With a dataset from a different release, run the
Cross Releasenotebook. Make sure to convert the dataset to the appropriate feature space and encoding with the providedretrofitandpreprocess\_and\_encodefunctions
The repository contains data gathered during 'Season of Plunder', 'Season of the Seraph' and 'Lightfall' respectively. It is present in the data directory, and available in raw and csv formats. Raw formats are to be processed by the Preprocess notebook again to account for changes in the game's database definitions.
Register an Application at Bungie's developer portal to obtain your API key