https://github.com/KDD-OpenSource/data-generation

The repository provides a synthetic multivariate time series data generator. The implementation is an extention of the cylinder-bell-funnel time series data generator. The scipt enables synthetic data generation of different length, dimensions and samples.
https://github.com/KDD-OpenSource/data-generation

multivariate-timeseries synthetic-data timeseries-data timeseriesclassification

Last synced: about 2 months ago
JSON representation

The repository provides a synthetic multivariate time series data generator. The implementation is an extention of the cylinder-bell-funnel time series data generator. The scipt enables synthetic data generation of different length, dimensions and samples.

• Host: GitHub
• URL: https://github.com/KDD-OpenSource/data-generation
• Owner: KDD-OpenSource
• Created: 2017-03-01T14:49:24.000Z (over 7 years ago)
• Default Branch: master
• Last Pushed: 2018-03-12T09:57:53.000Z (over 6 years ago)
• Last Synced: 2024-04-08T02:28:40.116Z (5 months ago)
• Topics: multivariate-timeseries, synthetic-data, timeseries-data, timeseriesclassification
• Language: Jupyter Notebook
• Homepage:
• Size: 789 KB
• Stars: 11
• Watchers: 4
• Forks: 4
• Open Issues: 1
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

• awesome-data-synthesis - data-generator

README

        
# data-generation

The repository provides a synthetic multivariate time series data generator. 

The implementation is an extension of the cylinder-bell-funnel [1] time series data generator. 

The script enables synthetic data generation of different length, dimensions and samples. 

Scaleability_length.ipynb --> Generate time series of different length and fixed number of samples/dimensions. 

Scaleability_number.ipynb --> Generate time series of different number of samples and fixed number of dimensions/length. 

Scaleability_robustness.ipynb --> Generate time series of different number of relevant and noisy dimensions and fixed number of samples/length. 

[1] http://timeseriesclassification.com/description.php?Dataset=CBF