TZIGANE

A Bokeh based application to visualize and edit timeseries. It aims at creating higher-level and easy-to-use objects. It packs a number of predefined data visualizations templates, along with edition toolbars.

Objectives

• Annotate Time Series for supervised learning
• Subject algorithm results to visual inspections
• Explore large swaths of data quickly and efficiently
• Manual edits to improve customer satisfaction pending full automation
• Generally create interactive data visualization toolbox to support learning, communication and marketing

Content of the repository

• static : static objects for the browser visualization
• templates: templates for the visualization
• tzigane: objects' definitions.

How to run Tzigane

[$ git clone ] [$ cd tzigane]

• install all the dependencies (or make sureyou have bokeh version 0.12.9 with [$ pip install bokeh==0.12.9]): [$ pip install -r requirements.txt]
• run the application (will automatically load the environment and devices and open a new window) [$ python main.py]

HEROKU Deployment

To deploy with heroku:

• [$ heroku create ]
• commit your changes
• [git push heroku master]

Visible on browser (https://.herokuapp.com/) with:

• [heroku open]

To check the logs when there is an issue:

• [heroku logs (--tail)]