Param subobjects

Download this notebook from GitHub (right-click to download).


In [ ]:
import param
import numpy as np

This example demonstrates how to use the Param library to express nested hierarchies of classes whose parameters can be edited in a GUI, without tying those classes to Panel or any other GUI framework.

For this purpose we create a hierarchy of three classes that draw Bokeh plots. At the top level there is a ShapeViewer that allows selecting between different Shape classes. The Shape classes include a subobject controlling the Style (i.e. the color and line_width) of the drawn shapes.

In each case, param.depends is used to indicate which parameter that computation depends on, either a parameter of this object (as for radius below) or a parameter of a subobject (i.e., a parameter of one of this object's parameters, as for style.color below).

In [ ]:
from bokeh.plotting import figure

class Style(param.Parameterized):
    
    color = param.Color(default='#0f6f0f')
    
    line_width = param.Number(default=2, bounds=(0, 10))


class Shape(param.Parameterized):

    radius = param.Number(default=1, bounds=(0, 1))
    
    style = param.Parameter(precedence=3)

    def __init__(self, **params):
        if 'style' not in params:
            params['style'] = Style(name='Style')
        super(Shape, self).__init__(**params)
        self.figure = figure(x_range=(-1, 1), y_range=(-1, 1), width=400, height=400)
        self.renderer = self.figure.line(*self._get_coords())
        self._update_style()

    @param.depends('style.color', 'style.line_width', watch=True)
    def _update_style(self):
        self.renderer.glyph.line_color = self.style.color
        self.renderer.glyph.line_width = self.style.line_width

    def _get_coords(self):
        return [], []

    def view(self):
        return self.figure


class Circle(Shape):

    n = param.Integer(default=100, precedence=-1)
    
    def __init__(self, **params):
        super(Circle, self).__init__(**params)
        
    def _get_coords(self):
        angles = np.linspace(0, 2*np.pi, self.n+1)
        return (self.radius*np.sin(angles),
                self.radius*np.cos(angles))
        
    @param.depends('radius', watch=True)
    def update(self):
        xs, ys = self._get_coords()
        self.renderer.data_source.data.update({'x': xs, 'y': ys})
    
class NGon(Circle):

    n = param.Integer(default=3, bounds=(3, 10), precedence=1)

    @param.depends('radius', 'n', watch=True)
    def update(self):
        xs, ys = self._get_coords()
        self.renderer.data_source.data.update({'x': xs, 'y': ys})
        
        
shapes = [NGon(name='NGon'), Circle(name='Circle')]

Having defined our basic domain model (of shapes in this case), we can now make a generic viewer using Panel without requiring or encoding information about the underlying domain objects. Here, we define a view method that will be called whenever any of the possible parameters that a shape might have changes, without necessarily knowing what those are (as for shape.param below). That way, if someone adds a Line shape that has no n parameter but has orientation, the viewer should continue to work and be responsive. We can also depend on specific parameters (as for shape.radius) if we wish. Either way, the panel should then reactively update to each of the shape's parameters as they are changed in the browser:

In [ ]:
import panel as pn

pn.extension()

class ShapeViewer(param.Parameterized):
    
    shape = param.ObjectSelector(default=shapes[0], objects=shapes)
    
    @param.depends('shape', 'shape.param')
    def view(self):
        return self.shape.view()

    @param.depends('shape', 'shape.radius')
    def title(self):
        return '## %s (radius=%.1f)' % (type(self.shape).__name__, self.shape.radius)
    
    def panel(self):
        return pn.Column(self.title, self.view)
    
    
# Instantiate and display ShapeViewer
viewer = ShapeViewer()
subpanel = pn.Column()

pn.Row(
    pn.Column(pn.Param(viewer.param, expand_layout=subpanel), subpanel),
    viewer.panel(),
).servable()

Download this notebook from GitHub (right-click to download).