bokeh.events¶
Represent granular events that can be used to trigger callbacks.
Bokeh documents and applications are capable of supporting various kinds of interactions. These are often associated with events, such as mouse or touch events, interactive downsampling mode activation, widget or tool events, and others. The classes in this module represent these different events, so that callbacks can be attached and executed when they occur.
It is possible to respond to events with CustomJS callbacks, which will
function with or without a Bokeh server. This can be accomplished by passing
and event class, and a CustomJS model to the
js_on_event() method. When the CustomJS is
executed in the browser, its cb_obj argument will contain the concrete
event object that triggered the callback.
from bokeh.events import ButtonClick
from bokeh.models import Button, CustomJS
button = Button()
button.js_on_event(ButtonClick, CustomJS(code='console.log("JS:Click")'))
Alternatively it is possible to trigger Python code to run when events
happen, in the context of a Bokeh application running on a Bokeh server.
This can accomplished by passing an event class, and a callback function
to the the on_event() method. The callback should
accept a single argument event, which will be passed the concrete
event object that triggered the callback.
from bokeh.events import ButtonClick
from bokeh.models import Button
button = Button()
def callback(event):
print('Python:Click')
button.on_event(ButtonClick, callback)
Note
There is no throttling of events. Some events such as MouseMove
may trigger at a very high rate.
-
class
DoubleTap(model, sx=None, sy=None, x=None, y=None)[source]¶ Announce a double-tap or double-click event on a Bokeh plot.
-
class
Event(model)[source]¶ Base class for all Bokeh events.
This base class is not typically useful to instantiate on its own.
-
__init__(model)[source]¶ Create a new base event.
Parameters: model (Model) – a Bokeh model to register event callbacks on
-
classmethod
decode_json(dct)[source]¶ Custom JSON decoder for Events.
Can be used as the
object_hookargument ofjson.loadorjson.loads.Parameters: dct (dict) – a JSON dictionary to decode The dictionary should have keys event_nameandevent_valuesRaises: ValueError, if the event_name is unknown Examples
>>> import json >>> from bokeh.events import Event >>> data = '{"event_name": "pan", "event_values" : {"model_id": 1, "x": 10, "y": 20, "sx": 200, "sy": 37}}' >>> json.loads(data, object_hook=Event.decode_json) <bokeh.events.Pan object at 0x1040f84a8>
-
-
class
LODStart(model)[source]¶ Announce the start of “interactive level-of-detail” mode on a plot.
During interactive actions such as panning or zooming, Bokeh can optionally, temporarily draw a reduced set of the data, in order to maintain high interactive rates. This is referred to as interactive Level-of-Detail (LOD) mode. This event fires whenever a LOD mode has just begun.
-
class
LODEnd(model)[source]¶ Announce the end of “interactive level-of-detail” mode on a plot.
During interactive actions such as panning or zooming, Bokeh can optionally, temporarily draw a reduced set of the data, in order to maintain high interactive rates. This is referred to as interactive Level-of-Detail (LOD) mode. This event fires whenever a LOD mode has just ended.
-
class
MouseEnter(model, sx=None, sy=None, x=None, y=None)[source]¶ Announce a mouse enter event onto a Bokeh plot.
Note
The enter event is generated when the mouse leaves the entire Plot canvas, including any border padding and space for axes or legends.
-
class
MouseLeave(model, sx=None, sy=None, x=None, y=None)[source]¶ Announce a mouse leave event from a Bokeh plot.
Note
The leave event is generated when the mouse leaves the entire Plot canvas, including any border padding and space for axes or legends.
-
class
MouseMove(model, sx=None, sy=None, x=None, y=None)[source]¶ Announce a mouse movement event over a Bokeh plot.
Note
This event can fire at a very high rate, potentially increasing network traffic or CPU load.
-
class
MouseWheel(model, delta=None, **kwargs)[source]¶ Announce a mouse wheel event on a Bokeh plot.
Note
By default, Bokeh plots do not prevent default scroll events unless a
WheelZoomToolorWheelPanToolis active. This may change in future releases.
-
class
Pan(model, delta_x=None, delta_y=None, direction=None, **kwargs)[source]¶ Announce a pan event on a Bokeh plot.
-
class
PanEnd(model, sx=None, sy=None, x=None, y=None)[source]¶ Announce the end of a pan event on a Bokeh plot.
-
class
PanStart(model, sx=None, sy=None, x=None, y=None)[source]¶ Announce the start of a pan event on a Bokeh plot.
-
class
Pinch(model, scale=None, **kwargs)[source]¶ Announce a pinch event on a Bokeh plot.
Note
This event is only applicable for touch-enabled devices.
-
class
PinchEnd(model, sx=None, sy=None, x=None, y=None)[source]¶ Announce the end of a pinch event on a Bokeh plot.
Note
This event is only applicable for touch-enabled devices.
-
class
PinchStart(model, sx=None, sy=None, x=None, y=None)[source]¶ Announce the start of a pinch event on a Bokeh plot.
Note
This event is only applicable for touch-enabled devices.
-
class
PointEvent(model, sx=None, sy=None, x=None, y=None)[source]¶ Base class for UI events associated with a specific (x,y) point.
Note that data space coordinates are relative to the default range, not any extra ranges, and the the screen space origin is at the top left of the HTML canvas.
-
class
Press(model, sx=None, sy=None, x=None, y=None)[source]¶ Announce a press event on a Bokeh plot.
-
class
SelectionGeometry(model, geometry=None, final=True)[source]¶ Announce the coordinates of a selection event on a plot.