On the chapter about Editable DataTable
we can find this :

This chapter includes recipes for:
    Reading the contents of the DataTable
    Filtering out null values
    Uploading data
    Determining which cell has changed 
    Adding or removing columns
    Adding or removing rows
    Clearable, deletable, renamable, and hideable colum

Most items coming from questions on the community forum

• a proper multicategory bar example based on https://community.plotly.com/t/multicategory-axis-type-example/26392
• example of shape or annotation covering several subplots thanks to xref='paper'. Also the shapes and annotations tutorial should link to each other.
• orthographic projection example in 3d axes tutorial (or

What likely needs to happen:

1. Docs moved inside of package (OR linked via setuptools)
2. Docs read and imported via python
3. Docs then parsed and available on a per-widget basis.

The ultimate goal of the above is so that there is never any disconnect between docs and UI. It should always stay in sync.

App name

dash-mapd-demo

Description of antipattern

Currently a new db connection is being made in every callback, and sometimes multiple times per callback. We should implement a connection pool (using sqlalchemy is probably easiest) and use that instead to improve performance and to demonstrate be