Adding A Custom Widget

This example shows how to add a double-ended slider widget to the plot.

The single normal Bokeh slider controls the power of the line. The double ended sliders control the x range for the line.

Python script:

"""Example implementation of two double ended sliders as extension widgets"""
from import Float, Instance, Tuple, Bool, Enum
from bokeh.models import InputWidget
from bokeh.models.callbacks import Callback
from bokeh.core.enums import SliderCallbackPolicy

from bokeh.layouts import column
from bokeh.models import Slider, CustomJS, ColumnDataSource
from import show
from bokeh.plotting import Figure

class IonRangeSlider(InputWidget):
    # The special class attribute ``__implementation__`` should contain a string
    # of JavaScript (or CoffeeScript) code that implements the JavaScript side
    # of the custom extension model or a string name of a JavaScript (or
    # CoffeeScript) file with the implementation.

    __implementation__ = ''
    __javascript__ = ["",
    __css__ = ["",

    # Below are all the "properties" for this model. Bokeh properties are
    # class attributes that define the fields (and their types) that can be
    # communicated automatically between Python and the browser. Properties
    # also support type validation. More information about properties in
    # can be found here:

    disable = Bool(default=True, help="""
    Enable or disable the slider.

    grid = Bool(default=True, help="""
    Show or hide the grid beneath the slider.

    start = Float(default=0, help="""
    The minimum allowable value.

    end = Float(default=1, help="""
    The maximum allowable value.

    range = Tuple(Float, Float, help="""
    The start and end values for the range.

    step = Float(default=0.1, help="""
    The step between consecutive values.

    callback = Instance(Callback, help="""
    A callback to run in the browser whenever the current Slider value changes.

    callback_throttle = Float(default=200, help="""
    Number of microseconds to pause between callback calls as the slider is moved.

    callback_policy = Enum(SliderCallbackPolicy, default="throttle", help="""
    When the callback is initiated. This parameter can take on only one of three options:
       "continuous": the callback will be executed immediately for each movement of the slider
       "throttle": the callback will be executed at most every ``callback_throttle`` milliseconds.
       "mouseup": the callback will be executed only once when the slider is released.
       The `mouseup` policy is intended for scenarios in which the callback is expensive in time.

x = [x*0.005 for x in range(2, 198)]
y = x

source = ColumnDataSource(data=dict(x=x, y=y))

plot = Figure(plot_width=400, plot_height=400)
plot.line('x', 'y', source=source, line_width=3, line_alpha=0.6, color='#ed5565')

callback_single = CustomJS(args=dict(source=source), code="""
        var data =;
        var f = cb_obj.value
        var x = data['x']
        var y = data['y']
        for (var i = 0; i < x.length; i++) {
            y[i] = Math.pow(x[i], f)

callback_ion = CustomJS(args=dict(source=source), code="""
        var data =;
        var f = cb_obj.range
        var x = data['x']
        var y = data['y']
        var pow = (Math.log(y[100])/Math.log(x[100]))
        var delta = (f[1] - f[0])/x.length
        for (var i = 0; i < x.length; i++) {
            x[i] = delta*i + f[0]
            y[i] = Math.pow(x[i], pow)

slider = Slider(start=0, end=5, step=0.1, value=1, title="Bokeh Slider - Power", callback=callback_single)
ion_range_slider = IonRangeSlider(start=0.01, end=0.99, step=0.01, range=(min(x), max(x)),
    title='Ion Range Slider - Range', callback=callback_ion, callback_policy='continuous')

layout = column(plot, slider, ion_range_slider)

CoffeeScript for ion range slider:

import {throttle} from "core/util/callback"

# The "core/properties" module has all the property types
import * as p from "core/properties"

# HTML construction and manipulation functions
import {empty, label, input, div} from "core/dom"

# We will subclass in JavaScript from the same class that was subclassed
# from in Python
import {InputWidget, InputWidgetView} from "models/widgets/input_widget"

# This model will actually need to render things, so we must provide
# view. The LayoutDOM model has a view already, so we will start with that
export class IonRangeSliderView extends InputWidgetView
  _createElement: () ->
    el = super()
    @$el = $(el)
    return el

  template: () ->
    if @model.title?
      if @model.title.length != 0
        title = label({for:}, " #{@model.title}: ")
      value = input({type: "text", id:, readonly: true})

    return (
      div({class: "bk-slider-parent"},
        div({class: "bk-slider-horizontal"},
          input({type: "text", class: "slider", id:})

  initialize: (options) ->
    @connect(@model.change, @render)
    @callbackWrapper = null
    if @model.callback_policy == 'continuous'
      @callbackWrapper = () ->
    if @model.callback_policy == 'throttle' and @model.callback
      @callbackWrapper = throttle(() ->
      , @model.callback_throttle)

  render: () ->
    # BokehJS Views create <div> elements by default, accessible as @$el.
    # Many Bokeh views ignore this default <div>, and instead do things
    # like draw to the HTML canvas. In this case though, we change the
    # contents of the <div>, based on the current slider value.
    # Set up parameters
    max = @model.end
    min = @model.start
    grid = @model.grid
    disable = @model.disabled
    range = @model.range or [max, min]
    [from, to] = range
    step = @model.step or ((max - min)/50)
    opts = {
      type: "double",
      grid: grid,
      min: min,
      max: max,
      from: from,
      to: to,
      step: step,
      disable: disable,
      onChange: (data) => @slide(data),
      onFinish: (data) => @slidestop(data),

    inp = @$el.find('.slider')[0]
    slider = jQuery(inp).ionRangeSlider(opts)
    range = [from, to]
    @$el.find( "##{ }" ).val( range.join(' - '))
    return @

  slidestop: (data) ->
    if @model.callback_policy == 'mouseup' or @model.callback_policy == 'throttle'

  slide: (data) ->
    range = [data.from,]
    value = range.join(' - ')
    @$el.find( "##{ }" ).val( value )
    @model.range = range
    if @callbackWrapper then @callbackWrapper()

export class IonRangeSlider extends InputWidget

  # If there is an associated view, this is boilerplate.
  default_view: IonRangeSliderView

  # The ``type`` class attribute should generally match exactly the name
  # of the corresponding Python class.
  type: "IonRangeSlider"

  # The @define block adds corresponding "properties" to the JS model. These
  # should basically line up 1-1 with the Python model class. Most property
  # types have counterparts, e.g. will be
  # p.String in the JS implementation. Where the JS type system is not yet
  # as rich, you can use p.Any as a "wildcard" property type.
  @define {
      range:             [ p.Any,                      ]
      start:             [ p.Number,      0            ]
      end:               [ p.Number,      1            ]
      step:              [ p.Number,      0.1          ]
      grid:              [ p.Bool,        true         ]
      callback_throttle: [ p.Number,      200          ]
      callback_policy:   [ p.String,      "throttle"   ]