A Python based framework to use ELKI's algorithms from the provided jar.
Faouzi Amrouche
What it does : Ignores commented files (by default with '@') and deletes label (by default last column) python pre-data.py -i -o (by default : inputfile = 'Shuttle_dataset.csv' and outputfile = 'Shuttle_dataset_ready.csv')
What it does : Runs the 11 algorithms ["COF","FastABOD","INFLO","KNN","KNNW","LDOF","LOF","LoOP","ODIN","SimplifiedLOF","KDEOS"] on the input data with variation of parameters (100 different values) and produces results under ./results/ python pre-data.py -i (by default : inputfile = 'Shuttle_dataset.csv')
What it does : Pick n times random results and creates result files under results folder. python concat.py -i