MTSClean

Description

This is the experiment code for the MTSClean: "Efficient Constraint-based Cleaning for Multi-Dimensional Time Series Data". In addition, it also includes some constraint mining algorithms for time series data, such as RFDiscover, DAFDiscover and tsdd

Requirements

• numpy
• pandas
• scipy
• deap
• pykalman
• multiprocessing
• tkinter
• tqdm

Running

1. Remove the timestamp from the data file and convert it to a csv file.
2. run

python ./experiments/comparison.py