fillplots API

class fillplots.Plotter(regions, config='default', xlim=(-10, 10), ylim=(-10, 10))

Plotter to fill regions with colors.

Parameters:

• regions (list) – Each value of list must be a “region specifier” (see below).
• config (str or Config) – 'default', 'none' or a Config object.
• xlim ((int, int)) – Limit for x-axis. Default is (-10, 10).
• ylim ((int, int)) – Limit for y-axis. Default is (-10, 10).

“Region specifier” is a list of “inequality specifiers”. Each “inequality specifier” is a tuple of (data[, less[, domain]]).

data is either a callable or a number. If it is a callable, this defines y of boundary as a function of x. If it is a number it is a vertical boundary, inidicated by that number.

less is a bool. True means < (less than) and False means > (larger than). Default is False.

domain is a pair (min, max) which indicates defined region of the function. It can be None, which means that this inequality is defined for any real number.

An example of specifying “x^2 > 0 and x + 5 > 0” is:

>>> plotter = Plotter(
...     [  # Regions:
...         [  # Inequalities:
...              (lambda x: x ** 2,),  # <-- Boundary data
...              (lambda x: x + 5,)
...         ],
...     ])

Internal representation

You can access internal representation as follows. First, regions attribute of this class holds a list of BaseRegion instances:

>>> plotter.regions
[<fillplots.regions... object at 0x...>]

Each region object has inequalities attribute, which holds a list of BaseInequality instances:

>>> plotter.regions[0].inequalities    
[<fillplots.inequalities... object at 0x...>,
 <fillplots.inequalities... object at 0x...>]

Finally, each inequality object has boundary attribute to hold an instance of BaseBoundary.

>>> plotter.regions[0].inequalities[0].boundary
<fillplots.boundaries... object at 0x...>

Configuration interface

Each of these “layer” has an instance of Config object whose attributes can be modified.

>>> plotter.regions[0].inequalities[0].boundary.config
<fillplots.core.Config object at 0x...>
>>> plotter.regions[0].inequalities[0].config
<fillplots.core.Config object at 0x...>
>>> plotter.regions[0].config
<fillplots.core.Config object at 0x...>
>>> plotter.config
<fillplots.core.Config object at 0x...>

Modifying upstream configuration propagates to downstream configuration. Let’s consider the following configuration:

>>> plotter.config.line_args['ls'] = 'dotted'
>>> plotter.config.line_args['lw'] = 8
>>> plotter.regions[0].inequalities[0].config.line_args['ls'] = 'dashed'

The boundary object of the 0th inequality mixes the configurations of the inequality and the root plotter object.

>>> plotter.regions[0].inequalities[0].boundary.config.line_args['ls']
'dashed'
>>> plotter.regions[0].inequalities[0].boundary.config.line_args['lw']
8

The configuration for the 0th inequality does not effect to the 1st inequality and its down stream. So, the configuration of the root plotter object is used:

>>> plotter.regions[0].inequalities[1].boundary.config.line_args['ls']
'dotted'

config

An instance of Config.

You can’t set this attribute (i.e., self.config = config raises an error). Use config._set_base instead.

ax

matplotlib.axes.Axes instance used by this plotter.

plot()

Plot regions and boundaries.

plot_boundaries()

Plot boundaries.

plot_positive_direction()

Plot direction that makes LHS of the inequality positive.

plot_regions()

Plot regions.

fillplots.plot_regions(regions, *args, **kwds)

Create Plotter object and call plot function of it.

All arguments are passed to Plotter.

Return type:Plotter

fillplots.boundary(function_or_number, domain=None)

Boundary factory function.

Parameters:

• function_or_number (callable or number) – If it is a callable, it is assumed to be a function that maps x to y. If it is a number, the boundary is a straight line specified by x = <number>.
• domain ((number, number) or None) – The boundary is defined on this domain. If it is None, the boundary is defined for any real number. If the argument function_or_number is a callable, the domain is on x-axis. If function_or_number is a number, the domain is on y-axis.

Return type:

BaseBoundary

Returns:

An instance of BaseBoundary subclass.

fillplots.annotate_regions(regions, text, horizontalalignment='center', verticalalignment='center', **kwds)

Annotate regions with text.

Put only one annotation for each contiguous group of regions.

Parameters:

• regions (list of BaseRegion) – These regions are annotated with the same text.
• text (str) – Annotation text.

Other keywords are passed to matplotlib.axes.Axes.text().

Internal classes

These classes are not meant to initialized from outside of this library. But you can access their instance via Plotter and call their methods.

class fillplots.core.Configurable(baseconfig)

config

An instance of Config.

You can’t set this attribute (i.e., self.config = config raises an error). Use config._set_base instead.

ax

matplotlib.axes.Axes instance used by this plotter.

class fillplots.boundaries.BaseBoundary(config, domain=None)

plot_boundary()

Plot this boundary.

class fillplots.inequalities.BaseInequality(baseconfig, data, less=False, domain=None)

boundary = None

An instance of BaseBoundary instance.

plot_positive_direction()

Plot direction that makes LHS of the inequality positive.

class fillplots.regions.BaseRegion(config, ineqs)

inequalities = None

List of BaseInequality instances.

plot_boundaries()

Plot boundaries.

plot_region()

Plot this region.

Configuration interface

class fillplots.core.Config(*args, **kwds)

Configuration interface.

See also Struct, which is a parent class of this class.

line_args = None

Arguments passed to matplotlib.axes.Axes.plot() or matplotlib.axes.Axes.axvline().

Default is {}.

fill_args = None

Arguments passed to matplotlib.axes.Axes.fill_between().

It can take additional keyword argument autocolor, which is not defined in matplotlib. autocolor can be used to set facecolor and edgecolor to the same color. If facecolor is specified, it is used. Otherwise, the color is generated from fill_color_cycle. The reason to add this argument is because it looks like that there is no good way to eliminate edge color natively by matplotlib [1].

Default is {}.

[1]	See: http://stackoverflow.com/questions/14143092/ http://permalink.gmane.org/gmane.comp.python.matplotlib.general/996

num_boundary_samples = None

Number of points to be used to draw boundaries.

Default is 1000.

num_com_samples = None

Number of points to be used to estimate center of mass of region.

Default is 50.

num_direction_arrows = None

Number of arrows per boundary to indicate positive directions.

Default is 5. It is used by fillplots.api.Plotter.plot_positive_direction() etc.

direction_arrows_size = None

Size of direction arrow as a ratio to axis length.

Default is 0.03. See also num_direction_arrows.

class fillplots.utils.chainstruct.Struct(*args, **kwds)

Chain-able struct (Like ChainMap, but for dot-access).

>>> zero = Struct()       # most upstream struct
>>> one = Struct(zero)
>>> two = Struct(one)     # most downstream struct

Setting attributes of the upstream objects “propagate” to downstream objects.

>>> zero.alpha = 0
>>> two.alpha
0
>>> one.alpha = 1
>>> (zero.alpha, one.alpha, two.alpha)
(0, 1, 1)

Struct has special treatment for dictionary attributes. Dictionaries of the same attribute in the upstream structs are mixed in the down stream structs.

>>> zero.beta = {'gamma': 10}
>>> one.beta = {'delta': 20}
>>> two.beta['gamma']
10
>>> two.beta['delta']
20
>>> two.beta == {'gamma': 10, 'delta': 20}
True

Note that changing downstream dictionaries does not change upstream ones.

>>> 'delta' in zero.beta
False
>>> two.beta['epsilon'] = 30
>>> 'epsilon' in one.beta
False

Setting whole dictionary works as expected.

>>> one.beta = {'gamma': 110, 'delta': 120}
>>> two.beta['gamma']
110
>>> two.beta['delta']
120