Is the Type "Datatime" or "Datetime"? (Inconsistent documentation)

Dash Python
1

I’m using types and formats for the first time with Dashtable and I noticed a discrepancy in the documentation.

In the typing page is states:

  • datatime : string in the form ‘YYYY-MM-DD HH:MM:SS.ssssss’ or some truncation thereof

But in the table reference (and several examples) it states:

type (one of: ‘any’, ‘numeric’, ‘text’, ‘datetime’)

What’s the correct one?

2

Also I don’t mean to hark on about this but the table reference is close to unusable, the only reason I noticed this was because of one of the examples.

Read the below for the columns attributes:

Attribute:

columns

Description:

Columns describes various aspects about each individual column. name and id are the only required parameters.

Type:

list of dict containing key(s): ‘options’, ‘type’, ‘format’, ‘editable’, ‘hidden’, ‘deletable’, ‘id’, ‘presentation’, ‘editable_name’, ‘Type’, ‘name’, ‘validation’, ‘on_change’, ‘clearable’

. Those keys have the following types:

  • options: . options has the following type: list of dict containing key(s): ‘label’, ‘value’

. Those keys have the following types:

  • label (number | string)
  • value (number | string)
  • type (one of: ‘any’, ‘numeric’, ‘text’, ‘datetime’)
  • format: . format has the following type: dict containing key(s): ‘specifier’, ‘locale’, ‘prefix’, ‘nully’

. Those keys have the following types:

  • specifier (string)
  • locale: . locale has the following type: dict containing key(s): ‘numerals’, ‘separate_4digits’, ‘decimal’, ‘percent’, ‘symbol’, ‘grouping’, ‘group’

. Those keys have the following types:

  • numerals (list of string)
  • separate_4digits (boolean)
  • decimal (string)
  • percent (string)
  • symbol (list of string)
  • grouping (list of number)
  • group (string)
  • prefix (number)
  • nully (boolean | number | component_name | dict | list)
  • editable (boolean)
  • hidden (boolean)
  • deletable (boolean | number)
  • id (string)
  • presentation (one of: ‘input’, ‘dropdown’)
  • editable_name (boolean | number)
  • Type ()
  • name (string | list of string)
  • validation: . validation has the following type: dict containing key(s): ‘allow_YY’, ‘default’, ‘allow_null’

. Those keys have the following types:

  • allow_YY (boolean)
  • default (boolean | number | component_name | dict | list)
  • allow_null (boolean)
  • on_change: . on_change has the following type: dict containing key(s): ‘failure’, ‘action’

. Those keys have the following types:

  • failure (one of: ‘accept’, ‘default’, ‘reject’)
  • action (one of: ‘coerce’, ‘none’, ‘validate’)
  • clearable (boolean)

Default Value:

There are 14 keys described at the beginning and 6 “Those keys have the following types” statements with no reference to which key they are describing. I think some of this is a formatting issue but I can’t tell.

3

Yeah, sorry about this. This is an issue with our auto-generating docstrings. It doesn’t handle deeply nested properties very well right now.

You may find it easier to read the annotated properties in the source code: dash-table/src/dash-table/dash/DataTable.js at a410e0969a710555e6f3f57c4df32216e56120f8 · plotly/dash-table · GitHub

These comments and property definitions are in React.js, but the Python classes (and documentation) are autogenerated from that exact definition.

2 Likes
4

Thanks, looks like the answer is datetime:

 type: PropTypes.oneOf(['any', 'numeric', 'text', 'datetime'])

If you haven’t already I’ll file a bug when I got home against Dash Docs for using datatime in Typing and User Input Processing | Dash for Python Documentation | Plotly