Interact with a database of tidy data tables using Bokeh

The purpose of this 'bokeh serve' example is to give a template for visualizing typical measurement databases.

If using the "tidy data" format, a database should be comprising of several tables, each representing a single observational unit (e.g. a sample test in certain experimental conditions). This unit comprises of variables (the columns) recorded in successive observations (the rows). See Wickham's paper

For vizualizing purpose, it is practical to have a dashboard with a main tab or window listing a summary of the information contained in each table. Then select one or several of them and plot/visualize their content. The objective is to be able to compare efficiently variables against each other, but also observational units against each other.

This template produces random csv files in a tests/ folder the first time you start it.

Some functionalities of this template: - list csv files and their info in a main tab - plot the content of a selected csv file, selecting x-axis, y-axis and optionaly a secondary y-axis - download in Excel format the transformed table (javascript implementation) - status text

You need python 3.6 and to install bokeh, pandas and all their dependencies.

If you use Anaconda, you can also go to the softfocus folder and create a new environment from there:

conda env create -f environment.yml

or make sure your current environment has the necessary packages

conda env update -f environment.yml

Simply download the folder.

softfocus uses the "folder" version of bokeh server.

Open a CLI, cd to the top softfocus folder then write:

bokeh serve softfocus

optional arguments:

    --allow-websocket-origin=localhost:5006 \\local access
    --allow-websocket-origin=REMOTE_IP:5006 \\remote access
    --show \\immediately opens a browser tab with the bokeh app
    --args folder/  \\list csv files from designated folder

The first time you run it, a sample data set will be generated in tests/ if you haven't done so yet.

Things to add/improve in the template:

• delete excess xls files in a separate thread without document lock
• use Tornado (or anything else) to implement the download method without dummy
• make a main class for the main tab/multi tab functionality, then subclasses for the type of database (csv folders, sql, hd5...)
• add SQL functionality
• filter or use a custom script on a table

Hugues-Yanis Amanieu, hyamani.eu