(september 7th 2018) https://dash.plot.ly/getting-started-part-2
Dash is a library written in python used to create reactive (think responsive like excel files) web applications. It also provides the ability to create complex data visualizations - given that it is built by the developers of plot.ly.
This set of python files are examples from the official tutorials that explain "getting started with Dash". It steps form a simple rendering of a static page through to an interactive visualization using plot.ly without user inputs/ dynamic updating all the way to dynamic pages interactions based on sliders/ other input component types.
My original thought was to play around and figure out how to build two project UI's:
- A demo visualization for a recommendation engine - originally for item-item reco based on selecting some item of interest, dynamically serving option reco's and displaying them visually
- A GUI for a clustering tool. Giving users the ability to select attributes they are interested in focusing on for analysis then on the backend running a pipeline that:
- decorrelates variables
- runs different, selected clustering algorithms
- reduces dimensionality and visualises the results
- produces a table with units assigned to clusters for consumption by user (perhaps ablity to download a csv/ excel/write to db with cluster profiles / assignments)
- produce a profile of the clusters based on other attributes pre-specified and the variables used to cluster