bokeh.layouts

Functions for arranging bokeh layout objects.

class GridSpec(nrows, ncols)[source]

Simplifies grid layout specification.

class Spacer(*args, **kwargs)[source]

Bases: bokeh.models.layouts.LayoutDOM

A container for space used to fill an empty spot in a row or column.

align
Property type

Either(Enum(Align), Tuple(Enum(Align), Enum(Align)))

Default value

'start'

The alignment point within the parent container.

This property is useful only if this component is a child element of a layout (e.g. a grid). Self alignment can be overridden by the parent container (e.g. grid track align).

aspect_ratio
Property type

Either(Enum(Enumeration(auto)), Float)

Default value

None

Describes the proportional relationship between component’s width and height.

This works if any of component’s dimensions are flexible in size. If set to a number, width / height = aspect_ratio relationship will be maintained. Otherwise, if set to "auto", component’s preferred width and height will be used to determine the aspect (if not set, no aspect will be preserved).

background
Property type

Color

Default value

None

Background color of the component.

Acceptable values are:

  • any of the named CSS colors, e.g 'green', 'indigo'

  • RGB(A) hex strings, e.g., '#FF0000', '#44444444'

  • CSS4 color strings, e.g., 'rgba(255, 0, 127, 0.6)', 'rgb(0 127 0 / 1.0)'

  • a 3-tuple of integers (r, g, b) between 0 and 255

  • a 4-tuple of (r, g, b, a) where r, g, b are integers between 0..255 and a is between 0..1

  • a 32-bit unsiged integers using the 0xRRGGBBAA byte order pattern

css_classes
Property type

List(String)

Default value

[]

A list of CSS class names to add to this DOM element. Note: the class names are simply added as-is, no other guarantees are provided.

It is also permissible to assign from tuples, however these are adapted – the property will always contain a list.

disabled
Property type

Bool

Default value

False

Whether the widget will be disabled when rendered.

If True, the widget will be greyed-out and not responsive to UI events.

height
Property type

NonNegativeInt

Default value

None

The height of the component (in pixels).

This can be either fixed or preferred height, depending on height sizing policy.

height_policy
Property type

Either(Auto, Enum(SizingPolicy))

Default value

'auto'

Describes how the component should maintain its height.

"auto"

Use component’s preferred sizing policy.

"fixed"

Use exactly height pixels. Component will overflow if it can’t fit in the available vertical space.

"fit"

Use component’s preferred height (if set) and allow to fit into the available vertical space within the minimum and maximum height bounds (if set). Component’s height neither will be aggressively minimized nor maximized.

"min"

Use as little vertical space as possible, not less than the minimum height (if set). The starting point is the preferred height (if set). The height of the component may shrink or grow depending on the parent layout, aspect management and other factors.

"max"

Use as much vertical space as possible, not more than the maximum height (if set). The starting point is the preferred height (if set). The height of the component may shrink or grow depending on the parent layout, aspect management and other factors.

Note

This is an experimental feature and may change in future. Use it at your own discretion. Prefer using sizing_mode if this level of control isn’t strictly necessary.

js_event_callbacks
Property type

Dict(String, List(Instance(CustomJS)))

Default value

{}

A mapping of event names to lists of CustomJS callbacks.

Typically, rather then modifying this property directly, callbacks should be added using the Model.js_on_event method:

callback = CustomJS(code="console.log('tap event occurred')")
plot.js_on_event('tap', callback)
js_property_callbacks
Property type

Dict(String, List(Instance(CustomJS)))

Default value

{}

A mapping of attribute names to lists of CustomJS callbacks, to be set up on BokehJS side when the document is created.

Typically, rather then modifying this property directly, callbacks should be added using the Model.js_on_change method:

callback = CustomJS(code="console.log('stuff')")
plot.x_range.js_on_change('start', callback)
margin
Property type

Tuple(Int, Int, Int, Int)

Default value

(0, 0, 0, 0)

Allows to create additional space around the component. The values in the tuple are ordered as follows - Margin-Top, Margin-Right, Margin-Bottom and Margin-Left, similar to CSS standards. Negative margin values may be used to shrink the space from any direction.

max_height
Property type

NonNegativeInt

Default value

None

Maximal height of the component (in pixels) if height is adjustable.

max_width
Property type

NonNegativeInt

Default value

None

Maximal width of the component (in pixels) if width is adjustable.

min_height
Property type

NonNegativeInt

Default value

None

Minimal height of the component (in pixels) if height is adjustable.

min_width
Property type

NonNegativeInt

Default value

None

Minimal width of the component (in pixels) if width is adjustable.

name
Property type

String

Default value

None

An arbitrary, user-supplied name for this model.

This name can be useful when querying the document to retrieve specific Bokeh models.

>>> plot.circle([1,2,3], [4,5,6], name="temp")
>>> plot.select(name="temp")
[GlyphRenderer(id='399d53f5-73e9-44d9-9527-544b761c7705', ...)]

Note

No uniqueness guarantees or other conditions are enforced on any names that are provided, nor is the name used directly by Bokeh for any reason.

sizing_mode
Property type

Enum(SizingMode)

Default value

None

How the component should size itself.

This is a high-level setting for maintaining width and height of the component. To gain more fine grained control over sizing, use width_policy, height_policy and aspect_ratio instead (those take precedence over sizing_mode).

Possible scenarios:

"fixed"

Component is not responsive. It will retain its original width and height regardless of any subsequent browser window resize events.

"stretch_width"

Component will responsively resize to stretch to the available width, without maintaining any aspect ratio. The height of the component depends on the type of the component and may be fixed or fit to component’s contents.

"stretch_height"

Component will responsively resize to stretch to the available height, without maintaining any aspect ratio. The width of the component depends on the type of the component and may be fixed or fit to component’s contents.

"stretch_both"

Component is completely responsive, independently in width and height, and will occupy all the available horizontal and vertical space, even if this changes the aspect ratio of the component.

"scale_width"

Component will responsively resize to stretch to the available width, while maintaining the original or provided aspect ratio.

"scale_height"

Component will responsively resize to stretch to the available height, while maintaining the original or provided aspect ratio.

"scale_both"

Component will responsively resize to both the available width and height, while maintaining the original or provided aspect ratio.

subscribed_events
Property type

List(String)

Default value

[]

List of events that are subscribed to by Python callbacks. This is the set of events that will be communicated from BokehJS back to Python for this model.

syncable
Property type

Bool

Default value

True

Indicates whether this model should be synchronized back to a Bokeh server when updated in a web browser. Setting to False may be useful to reduce network traffic when dealing with frequently updated objects whose updated values we don’t need.

Note

Setting this property to False will prevent any on_change() callbacks on this object from triggering. However, any JS-side callbacks will still work.

tags
Property type

List(AnyRef)

Default value

[]

An optional list of arbitrary, user-supplied values to attach to this model.

This data can be useful when querying the document to retrieve specific Bokeh models:

>>> r = plot.circle([1,2,3], [4,5,6])
>>> r.tags = ["foo", 10]
>>> plot.select(tags=['foo', 10])
[GlyphRenderer(id='1de4c3df-a83d-480a-899b-fb263d3d5dd9', ...)]

Or simply a convenient way to attach any necessary metadata to a model that can be accessed by CustomJS callbacks, etc.

Note

No uniqueness guarantees or other conditions are enforced on any tags that are provided, nor are the tags used directly by Bokeh for any reason.

visible
Property type

Bool

Default value

True

Whether the component will be visible and a part of a layout.

width
Property type

NonNegativeInt

Default value

None

The width of the component (in pixels).

This can be either fixed or preferred width, depending on width sizing policy.

width_policy
Property type

Either(Auto, Enum(SizingPolicy))

Default value

'auto'

Describes how the component should maintain its width.

"auto"

Use component’s preferred sizing policy.

"fixed"

Use exactly width pixels. Component will overflow if it can’t fit in the available horizontal space.

"fit"

Use component’s preferred width (if set) and allow it to fit into the available horizontal space within the minimum and maximum width bounds (if set). Component’s width neither will be aggressively minimized nor maximized.

"min"

Use as little horizontal space as possible, not less than the minimum width (if set). The starting point is the preferred width (if set). The width of the component may shrink or grow depending on the parent layout, aspect management and other factors.

"max"

Use as much horizontal space as possible, not more than the maximum width (if set). The starting point is the preferred width (if set). The width of the component may shrink or grow depending on the parent layout, aspect management and other factors.

Note

This is an experimental feature and may change in future. Use it at your own discretion. Prefer using sizing_mode if this level of control isn’t strictly necessary.

apply_theme(property_values)

Apply a set of theme values which will be used rather than defaults, but will not override application-set values.

The passed-in dictionary may be kept around as-is and shared with other instances to save memory (so neither the caller nor the HasProps instance should modify it).

Parameters

property_values (dict) – theme values to use in place of defaults

Returns

None

classmethod dataspecs()

Collect the names of all DataSpec properties on this class.

This method always traverses the class hierarchy and includes properties defined on any parent classes.

Returns

names of DataSpec properties

Return type

set[str]

classmethod dataspecs_with_props()

Collect a dict mapping the names of all DataSpec properties on this class to the associated properties.

This method always traverses the class hierarchy and includes properties defined on any parent classes.

Returns

mapping of names and DataSpec properties

Return type

dict[str, DataSpec]

equals(other)

Structural equality of models.

Parameters

other (HasProps) – the other instance to compare to

Returns

True, if properties are structurally equal, otherwise False

Link two Bokeh model properties using JavaScript.

This is a convenience method that simplifies adding a CustomJS callback to update one Bokeh model property whenever another changes value.

Parameters
  • attr (str) – The name of a Bokeh property on this model

  • other (Model) – A Bokeh model to link to self.attr

  • other_attr (str) – The property on other to link together

  • attr_selector (Union[int, str]) – The index to link an item in a subscriptable attr

Added in version 1.1

Raises

ValueError

Examples

This code with js_link:

select.js_link('value', plot, 'sizing_mode')

is equivalent to the following:

from bokeh.models import CustomJS
select.js_on_change('value',
    CustomJS(args=dict(other=plot),
             code="other.sizing_mode = this.value"
    )
)

Additionally, to use attr_selector to attach the left side of a range slider to a plot’s x_range:

range_slider.js_link('value', plot.x_range, 'start', attr_selector=0)

which is equivalent to:

from bokeh.models import CustomJS
range_slider.js_on_change('value',
    CustomJS(args=dict(other=plot.x_range),
             code="other.start = this.value[0]"
    )
)
js_on_change(event, *callbacks)

Attach a CustomJS callback to an arbitrary BokehJS model event.

On the BokehJS side, change events for model properties have the form "change:property_name". As a convenience, if the event name passed to this method is also the name of a property on the model, then it will be prefixed with "change:" automatically:

# these two are equivalent
source.js_on_change('data', callback)
source.js_on_change('change:data', callback)

However, there are other kinds of events that can be useful to respond to, in addition to property change events. For example to run a callback whenever data is streamed to a ColumnDataSource, use the "stream" event on the source:

source.js_on_change('streaming', callback)
layout(side, plot)
classmethod lookup(name)

Find the PropertyDescriptor for a Bokeh property on a class, given the property name.

Parameters

name (str) – name of the property to search for

Returns

descriptor for property named name

Return type

PropertyDescriptor

on_change(attr, *callbacks)

Add a callback on this object to trigger when attr changes.

Parameters
  • attr (str) – an attribute name on this object

  • *callbacks (callable) – callback functions to register

Returns

None

Example:

widget.on_change('value', callback1, callback2, ..., callback_n)
on_event(event, *callbacks)

Run callbacks when the specified event occurs on this Model

Not all Events are supported for all Models. See specific Events in bokeh.events for more information on which Models are able to trigger them.

classmethod properties(with_bases=True)

Collect the names of properties on this class.

This method optionally traverses the class hierarchy and includes properties defined on any parent classes.

Parameters

with_bases (bool, optional) – Whether to include properties defined on parent classes in the results. (default: True)

Returns

property names

Return type

set[str]

classmethod properties_containers()

Collect the names of all container properties on this class.

This method always traverses the class hierarchy and includes properties defined on any parent classes.

Returns

names of container properties

Return type

set[str]

classmethod properties_with_refs()

Collect the names of all properties on this class that also have references.

This method always traverses the class hierarchy and includes properties defined on any parent classes.

Returns

names of properties that have references

Return type

set[str]

properties_with_values(include_defaults: bool = True) → Dict[str, Any]

Collect a dict mapping property names to their values.

This method always traverses the class hierarchy and includes properties defined on any parent classes.

Non-serializable properties are skipped and property values are in “serialized” format which may be slightly different from the values you would normally read from the properties; the intent of this method is to return the information needed to losslessly reconstitute the object instance.

Parameters

include_defaults (bool, optional) – Whether to include properties that haven’t been explicitly set since the object was created. (default: True)

Returns

mapping from property names to their values

Return type

dict

query_properties_with_values(query, include_defaults=True)

Query the properties values of HasProps instances with a predicate.

Parameters
  • query (callable) – A callable that accepts property descriptors and returns True or False

  • include_defaults (bool, optional) – Whether to include properties that have not been explicitly set by a user (default: True)

Returns

mapping of property names and values for matching properties

Return type

dict

references()

Returns all Models that this object has references to.

remove_on_change(attr, *callbacks)

Remove a callback from this object

select(selector)

Query this object and all of its references for objects that match the given selector.

Parameters

selector (JSON-like) –

Returns

seq[Model]

select_one(selector)

Query this object and all of its references for objects that match the given selector. Raises an error if more than one object is found. Returns single matching object, or None if nothing is found :param selector: :type selector: JSON-like

Returns

Model

set_from_json(name, json, models=None, setter=None)

Set a property value on this object from JSON.

Parameters
  • name – (str) : name of the attribute to set

  • json – (JSON-value) : value to set to the attribute to

  • models (dict or None, optional) –

    Mapping of model ids to models (default: None)

    This is needed in cases where the attributes to update also have values that have references.

  • setter (ClientSession or ServerSession or None, optional) –

    This is used to prevent “boomerang” updates to Bokeh apps.

    In the context of a Bokeh server application, incoming updates to properties will be annotated with the session that is doing the updating. This value is propagated through any subsequent change notifications that the update triggers. The session can compare the event setter to itself, and suppress any updates that originate from itself.

Returns

None

set_select(selector, updates)

Update objects that match a given selector with the specified attribute/value updates.

Parameters
  • selector (JSON-like) –

  • updates (dict) –

Returns

None

themed_values()

Get any theme-provided overrides.

Results are returned as a dict from property name to value, or None if no theme overrides any values for this instance.

Returns

dict or None

to_json(include_defaults)

Returns a dictionary of the attributes of this object, containing only “JSON types” (string, number, boolean, none, dict, list).

References to other objects are serialized as “refs” (just the object ID and type info), so the deserializer will need to separately have the full attributes of those other objects.

There’s no corresponding from_json() because to deserialize an object is normally done in the context of a Document (since the Document can resolve references).

For most purposes it’s best to serialize and deserialize entire documents.

Parameters

include_defaults (bool) – whether to include attributes that haven’t been changed from the default

to_json_string(include_defaults)

Returns a JSON string encoding the attributes of this object.

References to other objects are serialized as references (just the object ID and type info), so the deserializer will need to separately have the full attributes of those other objects.

There’s no corresponding from_json_string() because to deserialize an object is normally done in the context of a Document (since the Document can resolve references).

For most purposes it’s best to serialize and deserialize entire documents.

Parameters

include_defaults (bool) – whether to include attributes that haven’t been changed from the default

trigger(attr, old, new, hint=None, setter=None)
unapply_theme()

Remove any themed values and restore defaults.

Returns

None

update(**kwargs)

Updates the object’s properties from the given keyword arguments.

Returns

None

Examples

The following are equivalent:

from bokeh.models import Range1d

r = Range1d

# set properties individually:
r.start = 10
r.end = 20

# update properties together:
r.update(start=10, end=20)
update_from_json(json_attributes, models=None, setter=None)

Updates the object’s properties from a JSON attributes dictionary.

Parameters
  • json_attributes – (JSON-dict) : attributes and values to update

  • models (dict or None, optional) –

    Mapping of model ids to models (default: None)

    This is needed in cases where the attributes to update also have values that have references.

  • setter (ClientSession or ServerSession or None, optional) –

    This is used to prevent “boomerang” updates to Bokeh apps.

    In the context of a Bokeh server application, incoming updates to properties will be annotated with the session that is doing the updating. This value is propagated through any subsequent change notifications that the update triggers. The session can compare the event setter to itself, and suppress any updates that originate from itself.

Returns

None

property document

The Document this model is attached to (can be None)

property struct

A Bokeh protocol “structure” of this model, i.e. a dict of the form:

{
    'type' : << view model name >>
    'id'   : << unique model id >>
}

Additionally there may be a subtype field if this model is a subtype.

JSON Prototype
{
  "align": "start",
  "aspect_ratio": null,
  "background": null,
  "css_classes": [],
  "disabled": false,
  "height": null,
  "height_policy": "auto",
  "id": "12033",
  "js_event_callbacks": {},
  "js_property_callbacks": {},
  "margin": [
    0,
    0,
    0,
    0
  ],
  "max_height": null,
  "max_width": null,
  "min_height": null,
  "min_width": null,
  "name": null,
  "sizing_mode": null,
  "subscribed_events": [],
  "syncable": true,
  "tags": [],
  "visible": true,
  "width": null,
  "width_policy": "auto"
}
column(*args, **kwargs)[source]

Create a column of Bokeh Layout objects. Forces all objects to have the same sizing_mode, which is required for complex layouts to work.

Parameters
  • children (list of LayoutDOM) – A list of instances for the column. Can be any of the following - Plot, Widget, Row, Column, ToolbarBox, Spacer.

  • sizing_mode ("fixed", "stretch_both", "scale_width", "scale_height", "scale_both") – How will the items in the layout resize to fill the available space. Default is "fixed". For more information on the different modes see sizing_mode description on LayoutDOM.

Returns

A column of LayoutDOM objects all with the same sizing_mode.

Return type

Column

Examples

>>> column(plot1, plot2)
>>> column(children=[widgets, plot], sizing_mode='stretch_both')
grid(children=[], sizing_mode=None, nrows=None, ncols=None)[source]

Conveniently create a grid of layoutable objects.

Grids are created by using GridBox model. This gives the most control over the layout of a grid, but is also tedious and may result in unreadable code in practical applications. grid() function remedies this by reducing the level of control, but in turn providing a more convenient API.

Supported patterns:

  1. Nested lists of layoutable objects. Assumes the top-level list represents a column and alternates between rows and columns in subsequent nesting levels. One can use None for padding purpose.

    >>> grid([p1, [[p2, p3], p4]])
    GridBox(children=[
        (p1, 0, 0, 1, 2),
        (p2, 1, 0, 1, 1),
        (p3, 2, 0, 1, 1),
        (p4, 1, 1, 2, 1),
    ])
    
  2. Nested Row and Column instances. Similar to the first pattern, just instead of using nested lists, it uses nested Row and Column models. This can be much more readable that the former. Note, however, that only models that don’t have sizing_mode set are used.

    >>> grid(column(p1, row(column(p2, p3), p4)))
    GridBox(children=[
        (p1, 0, 0, 1, 2),
        (p2, 1, 0, 1, 1),
        (p3, 2, 0, 1, 1),
        (p4, 1, 1, 2, 1),
    ])
    
  3. Flat list of layoutable objects. This requires nrows and/or ncols to be set. The input list will be rearranged into a 2D array accordingly. One can use None for padding purpose.

    >>> grid([p1, p2, p3, p4], ncols=2)
    GridBox(children=[
        (p1, 0, 0, 1, 1),
        (p2, 0, 1, 1, 1),
        (p3, 1, 0, 1, 1),
        (p4, 1, 1, 1, 1),
    ])
    
gridplot(children, sizing_mode=None, toolbar_location='above', ncols=None, plot_width=None, plot_height=None, toolbar_options=None, merge_tools=True)[source]

Create a grid of plots rendered on separate canvases.

The gridplot function builds a single toolbar for all the plots in the grid. gridplot is designed to layout a set of plots. For general grid layout, use the layout() function.

Parameters
  • children (list of lists of Plot) – An array of plots to display in a grid, given as a list of lists of Plot objects. To leave a position in the grid empty, pass None for that position in the children list. OR list of Plot if called with ncols. OR an instance of GridSpec.

  • sizing_mode ("fixed", "stretch_both", "scale_width", "scale_height", "scale_both") – How will the items in the layout resize to fill the available space. Default is "fixed". For more information on the different modes see sizing_mode description on LayoutDOM.

  • toolbar_location (above, below, left, right) – Where the toolbar will be located, with respect to the grid. Default is above. If set to None, no toolbar will be attached to the grid.

  • ncols (int, optional) – Specify the number of columns you would like in your grid. You must only pass an un-nested list of plots (as opposed to a list of lists of plots) when using ncols.

  • plot_width (int, optional) – The width you would like all your plots to be

  • plot_height (int, optional) – The height you would like all your plots to be.

  • toolbar_options (dict, optional) – A dictionary of options that will be used to construct the grid’s toolbar (an instance of ToolbarBox). If none is supplied, ToolbarBox’s defaults will be used.

  • merge_tools (True, False) – Combine tools from all child plots into a single toolbar.

Returns

A row or column containing the grid toolbar and the grid

of plots (depending on whether the toolbar is left/right or above/below. The grid is always a Column of Rows of plots.

Return type

Row or Column

Examples

>>> gridplot([[plot_1, plot_2], [plot_3, plot_4]])
>>> gridplot([plot_1, plot_2, plot_3, plot_4], ncols=2, plot_width=200, plot_height=100)
>>> gridplot(
        children=[[plot_1, plot_2], [None, plot_3]],
        toolbar_location='right'
        sizing_mode='fixed',
        toolbar_options=dict(logo='gray')
    )
layout(*args, **kwargs)[source]

Create a grid-based arrangement of Bokeh Layout objects.

Parameters
  • children (list of lists of LayoutDOM) – A list of lists of instances for a grid layout. Can be any of the following - Plot, Widget, Row, Column, ToolbarBox, Spacer.

  • sizing_mode ("fixed", "stretch_both", "scale_width", "scale_height", "scale_both") – How will the items in the layout resize to fill the available space. Default is "fixed". For more information on the different modes see sizing_mode description on LayoutDOM.

Returns

A column of Row layouts of the children, all with the same sizing_mode.

Return type

Column

Examples

>>> layout([[plot_1, plot_2], [plot_3, plot_4]])
>>> layout(
        children=[
            [widget_1, plot_1],
            [slider],
            [widget_2, plot_2, plot_3]
        ],
        sizing_mode='fixed',
    )
row(*args, **kwargs)[source]

Create a row of Bokeh Layout objects. Forces all objects to have the same sizing_mode, which is required for complex layouts to work.

Parameters
  • children (list of LayoutDOM) – A list of instances for the row. Can be any of the following - Plot, Widget, Row, Column, ToolbarBox, Spacer.

  • sizing_mode ("fixed", "stretch_both", "scale_width", "scale_height", "scale_both") – How will the items in the layout resize to fill the available space. Default is "fixed". For more information on the different modes see sizing_mode description on LayoutDOM.

Returns

A row of LayoutDOM objects all with the same sizing_mode.

Return type

Row

Examples

>>> row(plot1, plot2)
>>> row(children=[widgets, plot], sizing_mode='stretch_both')
widgetbox(*args, **kwargs)[source]

Create a column of bokeh widgets with predefined styling.

Parameters
  • children (list of Widget) – A list of widgets.

  • sizing_mode ("fixed", "stretch_both", "scale_width", "scale_height", "scale_both") – How will the items in the layout resize to fill the available space. Default is "fixed". For more information on the different modes see sizing_mode description on LayoutDOM.

Returns

A column layout of widget instances all with the same sizing_mode.

Return type

WidgetBox

Examples

>>> widgetbox([button, select])
>>> widgetbox(children=[slider], sizing_mode='scale_width')