Source code for holoviews.plotting.bokeh.widgets
from __future__ import absolute_import, division, unicode_literals
import math
import json
from functools import partial
import param
import numpy as np
from bokeh.models.widgets import Select, Slider, AutocompleteInput, TextInput, Div
from bokeh.layouts import widgetbox, row, column
from ...core import Store, NdMapping, OrderedDict
from ...core.util import (drop_streams, unique_array, isnumeric,
wrap_tuple_streams, unicode)
from ..renderer import MIME_TYPES
from ..widgets import NdWidget, SelectionWidget, ScrubberWidget
from .util import serialize_json
[docs]class BokehServerWidgets(param.Parameterized):
"""
BokehServerWidgets create bokeh widgets corresponding to all the
key dimensions found on a BokehPlot instance. It currently supports
to types of widgets sliders (which may be discrete or continuous)
and dropdown widgets letting you select non-numeric values.
"""
display_options = param.Dict(default={}, doc="""
Additional options controlling display options of the widgets.""")
editable = param.Boolean(default=False, doc="""
Whether the slider text fields should be editable. Disabled
by default for a more compact widget layout.""")
position = param.ObjectSelector(default='right',
objects=['right', 'left', 'above', 'below'])
sizing_mode = param.ObjectSelector(default='fixed',
objects=['fixed', 'stretch_both', 'scale_width',
'scale_height', 'scale_both'])
width = param.Integer(default=250, doc="""
Width of the widget box in pixels""")
basejs = param.String(default=None, precedence=-1, doc="""
Defines the local CSS file to be loaded for this widget.""")
extensionjs = param.String(default=None, precedence=-1, doc="""
Optional javascript extension file for a particular backend.""")
css = param.String(default=None, precedence=-1, doc="""
Defines the local CSS file to be loaded for this widget.""")
def __init__(self, plot, renderer=None, **params):
super(BokehServerWidgets, self).__init__(**params)
self.plot = plot
streams = []
for stream in plot.streams:
if any(k in plot.dimensions for k in stream.contents):
streams.append(stream)
self.dimensions, self.keys = drop_streams(streams,
plot.dimensions,
plot.keys)
if renderer is None:
backend = Store.current_backend
self.renderer = Store.renderers[backend]
else:
self.renderer = renderer
# Create mock NdMapping to hold the common dimensions and keys
self.mock_obj = NdMapping([(k, None) for k in self.keys],
kdims=list(self.dimensions))
self.widgets, self.lookups = self.get_widgets()
self.subplots = {}
if self.plot.renderer.mode == 'default':
self.attach_callbacks()
self.state = self.init_layout()
self._queue = []
self._active = False
if hasattr(self.plot.document, 'on_session_destroyed'):
self.plot.document.on_session_destroyed(self.plot._session_destroy)
[docs] @classmethod
def create_widget(self, dim, holomap=None, editable=False):
""""
Given a Dimension creates bokeh widgets to select along that
dimension. For numeric data a slider widget is created which
may be either discrete, if a holomap is supplied or the
Dimension.values are set, or a continuous widget for
DynamicMaps. If the slider is discrete the returned mapping
defines a mapping between values and labels making it possible
sync the two slider and label widgets. For non-numeric data
a simple dropdown selection widget is generated.
"""
label, mapping = None, None
if holomap is None:
if dim.values:
if dim.default is None:
default = dim.values[0]
elif dim.default not in dim.values:
raise ValueError("%s dimension default %r is not in dimension values: %s"
% (dim, dim.default, dim.values))
else:
default = dim.default
value = dim.values.index(default)
if all(isnumeric(v) for v in dim.values):
values = sorted(dim.values)
labels = [unicode(dim.pprint_value(v)) for v in values]
if editable:
label = AutocompleteInput(value=labels[value], completions=labels,
title=dim.pprint_label)
else:
label = Div(text='<b>%s</b>' % dim.pprint_value_string(labels[value]))
widget = Slider(value=value, start=0, end=len(dim.values)-1, title=None, step=1)
mapping = list(enumerate(zip(values, labels)))
else:
values = [(v, dim.pprint_value(v)) for v in dim.values]
widget = Select(title=dim.pprint_label, value=values[value][0],
options=values)
else:
start = dim.soft_range[0] if dim.soft_range[0] else dim.range[0]
end = dim.soft_range[1] if dim.soft_range[1] else dim.range[1]
dim_range = end - start
int_type = isinstance(dim.type, type) and issubclass(dim.type, int)
if dim.step is not None:
step = dim.step
elif isinstance(dim_range, int) or int_type:
step = 1
else:
step = 10**((round(math.log10(dim_range))-3))
if dim.default is None:
default = start
elif (dim.default < start or dim.default > end):
raise ValueError("%s dimension default %r is not in the provided range: %s"
% (dim, dim.default, (start, end)))
else:
default = dim.default
if editable:
label = TextInput(value=str(default), title=dim.pprint_label)
else:
label = Div(text='<b>%s</b>' % dim.pprint_value_string(default))
widget = Slider(value=default, start=start,
end=end, step=step, title=None)
else:
values = (dim.values if dim.values else
list(unique_array(holomap.dimension_values(dim.name))))
if dim.default is None:
default = values[0]
elif dim.default not in values:
raise ValueError("%s dimension default %r is not in dimension values: %s"
% (dim, dim.default, values))
else:
default = dim.default
if isinstance(values[0], np.datetime64) or isnumeric(values[0]):
values = sorted(values)
labels = [dim.pprint_value(v) for v in values]
value = values.index(default)
if editable:
label = AutocompleteInput(value=labels[value], completions=labels,
title=dim.pprint_label)
else:
label = Div(text='<b>%s</b>' % (dim.pprint_value_string(labels[value])))
widget = Slider(value=value, start=0, end=len(values)-1, title=None, step=1)
else:
labels = [dim.pprint_value(v) for v in values]
widget = Select(title=dim.pprint_label, value=default,
options=list(zip(values, labels)))
mapping = list(enumerate(zip(values, labels)))
return widget, label, mapping
[docs] def get_widgets(self):
"""
Creates a set of widgets representing the dimensions on the
plot object used to instantiate the widgets class.
"""
widgets = OrderedDict()
mappings = {}
for dim in self.mock_obj.kdims:
holomap = None if self.plot.dynamic else self.mock_obj
widget, label, mapping = self.create_widget(dim, holomap, self.editable)
if label is not None and not isinstance(label, Div):
label.on_change('value', partial(self.on_change, dim, 'label'))
widget.on_change('value', partial(self.on_change, dim, 'widget'))
widgets[dim.pprint_label] = (label, widget)
if mapping:
mappings[dim.pprint_label] = OrderedDict(mapping)
return widgets, mappings
def init_layout(self):
widgets = [widget for d in self.widgets.values()
for widget in d if widget]
wbox = widgetbox(widgets, width=self.width)
if self.position in ['right', 'below']:
plots = [self.plot.state, wbox]
else:
plots = [wbox, self.plot.state]
layout_fn = row if self.position in ['left', 'right'] else column
layout = layout_fn(plots, sizing_mode=self.sizing_mode)
return layout
def on_change(self, dim, widget_type, attr, old, new):
self._queue.append((dim, widget_type, attr, old, new))
if not self._active:
self.plot.document.add_timeout_callback(self.update, 50)
self._active = True
[docs] def update(self):
"""
Handle update events on bokeh server.
"""
if not self._queue:
return
dim, widget_type, attr, old, new = self._queue[-1]
self._queue = []
dim_label = dim.pprint_label
label, widget = self.widgets[dim_label]
if widget_type == 'label':
if isinstance(label, AutocompleteInput):
value = [new]
widget.value = value
else:
widget.value = float(new)
elif label:
lookups = self.lookups.get(dim_label)
if not self.editable:
if lookups:
new = lookups[widget.value][1]
label.text = '<b>%s</b>' % dim.pprint_value_string(new)
elif isinstance(label, AutocompleteInput):
text = lookups[new][1]
label.value = text
else:
label.value = dim.pprint_value(new)
key = []
for dim, (label, widget) in self.widgets.items():
lookups = self.lookups.get(dim)
if label and lookups:
val = lookups[widget.value][0]
else:
val = widget.value
key.append(val)
key = wrap_tuple_streams(tuple(key), self.plot.dimensions,
self.plot.streams)
self.plot.update(key)
self._active = False
class BokehWidget(NdWidget):
css = param.String(default='bokehwidgets.css', doc="""
Defines the local CSS file to be loaded for this widget.""")
extensionjs = param.String(default='bokehwidgets.js', doc="""
Optional javascript extension file for a particular backend.""")
def _get_data(self):
msg, metadata = self.renderer.components(self.plot, comm=False)
data = super(BokehWidget, self)._get_data()
return dict(data, init_html=msg['text/html'],
init_js=msg[MIME_TYPES['js']],
plot_id=self.plot.state._id)
def encode_frames(self, frames):
if self.export_json:
self.save_json(frames)
frames = {}
else:
frames = json.dumps(frames).replace('</', r'<\/')
return frames
def get_frames(self):
nframes = len(self.plot)
if self.embed:
self.plot.update(nframes-1)
frames = OrderedDict([(idx, self._plot_figure(idx))
for idx in range(nframes)])
else:
frames = {}
return self.encode_frames(frames)
def _plot_figure(self, idx, fig_format='json'):
"""
Returns the figure in html format on the
first call and
"""
self.plot.update(idx)
if self.embed:
patch = self.renderer.diff(self.plot, binary=False)
msg = serialize_json(dict(content=patch.content,
root=self.plot.state._id))
return msg
class BokehSelectionWidget(BokehWidget, SelectionWidget):
def _get_data(self):
if not self.plot.dynamic:
widgets, _, _ = self.get_widgets()
key = tuple(w['value'] for w in widgets)
self.plot.update(key)
return super(BokehSelectionWidget, self)._get_data()
class BokehScrubberWidget(BokehWidget, ScrubberWidget):
pass