Time Series Slicer for Energy Data

The script is a fork from ts2img. The slicing of time series was a bi-product of the trasfomation, so this scipt utilizes only the latter part.

For easier data handling NILMTK was used.

environment

Works with all datasets supported by NILMTK. Tested on:

Install Instructions

❗️ If possible, install on a Linux machine.

Clone this repository and change directory to conda:

    git clone https://github.com/jenkoj/ts2img && cd ts2img/conda

Create a new environment by running:

    conda env create --name ts2img --file=ts2img.yml

Activate the newly created environment:
```
    conda activate ts2img
```
Get hold of a dataset converted to NILMTK format or convert your dataset.
- Check if your dataset has supported converter link, if not you can write your own dataset converter by following instructions here.
- Convert your dataset by following notebook here.
Place converted dataset in datasets directory.
Set parameters.
Finally, run:
```
    ipython -c "%run converter.ipynb"
```

When adjusting the parameters start with iAWE. Since it is small, it is easy to handle.

Name		Name	Last commit message	Last commit date
Latest commit History 158 Commits
conda		conda
datasets		datasets
imgs		imgs
utils		utils
.gitignore		.gitignore
readme.md		readme.md
slicer.ipynb		slicer.ipynb