pygal/svg/charts/plot.py

# -*- coding: utf-8 -*-

"plot.py"

import sys
from itertools import izip, count, chain
from lxml import etree

from pygal.graph import Graph

from .util import float_range


def get_pairs(i):
    i = iter(i)
    while True:
        yield i.next(), i.next()

# I'm not sure how this is more beautiful than ugly.
if sys.version >= '3':
    def apply(func):
        return func()


class Plot(Graph):
    """=== For creating SVG plots of scalar data

    = Synopsis

      require 'SVG/Graph/Plot'

      # Data sets are x,y pairs
      # Note that multiple data sets can differ in length, and that the
      # data in the datasets needn't be in order; they will be ordered
      # by the plot along the X-axis.
      projection = [
        6, 11,    0, 5,   18, 7,   1, 11,   13, 9,   1, 2,   19, 0,   3, 13,
        7, 9
     ]
      actual = [
        0, 18,    8, 15,    9, 4,   18, 14,   10, 2,   11, 6,  14, 12,
        15, 6,   4, 17,   2, 12
     ]

      graph = SVG::Graph::Plot.new({
       :height => 500,
           :width => 300,
        :key => true,
        :scale_x_integers => true,
        :scale_y_integerrs => true,
     })

      graph.add_data({
       :data => projection
         :title => 'Projected',
     })

      graph.add_data({
       :data => actual,
         :title => 'Actual',
     })

      print graph.burn()

    = Description

    Produces a graph of scalar data.

    This object aims to allow you to easily create high quality
    SVG[http://www.w3c.org/tr/svg] scalar plots. You can either use the
    default style sheet or supply your own. Either way there are many options
    which can be configured to give you control over how the graph is
    generated - with or without a key, data elements at each point, title,
    subtitle etc.

    = Examples

    http://www.germane-software/repositories/public/SVG/test/plot.rb

    = Notes

    The default stylesheet handles upto 10 data sets, if you
    use more you must create your own stylesheet and add the
    additional settings for the extra data sets. You will know
    if you go over 10 data sets as they will have no style and
    be in black.

    Unlike the other types of charts, data sets must contain x,y pairs:

      [1, 2]    # A data set with 1 point: (1,2)
      [1,2, 5,6] # A data set with 2 points: (1,2) and (5,6)

    = See also

    * SVG::Graph::Graph
    * SVG::Graph::BarHorizontal
    * SVG::Graph::Bar
    * SVG::Graph::Line
    * SVG::Graph::Pie
    * SVG::Graph::TimeSeries

    == Author

    Sean E. Russell <serATgermaneHYPHENsoftwareDOTcom>

    Copyright 2004 Sean E. Russell
    This software is available under the Ruby license[LICENSE.txt]"""

    top_align = right_align = top_font = right_font = 1

    """Determines the scaling for the Y axis divisions.

      graph.scale_y_divisions = 0.5

    would cause the graph to attempt to generate labels stepped by 0.5; EG:
    0, 0.5, 1, 1.5, 2, ..."""
    scale_y_divisions = None
    "Make the X axis labels integers"
    scale_x_integers = False
    "Make the Y axis labels integers"
    scale_y_integers = False
    "Fill the area under the line"
    area_fill = False
    """Show a small circle on the graph where the line
    goes from one point to the next."""
    show_data_points = True
    "Indicate whether the lines should be drawn between points"
    draw_lines_between_points = True
    "Set the minimum value of the X axis"
    min_x_value = None
    "Set the minimum value of the Y axis"
    min_y_value = None
    "Set the maximum value of the X axis"
    max_x_value = None
    "Set the maximum value of the Y axis"
    max_y_value = None

    stacked = False

    stylesheet_names = Graph.stylesheet_names + ['plot.css']

    @apply
    def scale_x_divisions():
        doc = """Determines the scaling for the X axis divisions.

            graph.scale_x_divisions = 2

            would cause the graph to attempt
            to generate labels stepped by 2; EG:
            0,2,4,6,8..."""

        def fget(self):
            return getattr(self, '_scale_x_divisions', None)

        def fset(self, val):
            self._scale_x_divisions = val
        return property(**locals())

    def validate_data(self, data):
        if len(data['data']) % 2 != 0:
            raise ValueError(
                "Expecting x,y pairs for data points for %s." %
                self.__class__.__name__)

    def process_data(self, data):
        pairs = list(get_pairs(data['data']))
        pairs.sort()
        data['data'] = zip(*pairs)

    def calculate_left_margin(self):
        super(Plot, self).calculate_left_margin()
        label_left = len(str(
            self.get_x_labels()[0])) / 2 * self.font_size * 0.6
        self.border_left = max(label_left, self.border_left)

    def calculate_right_margin(self):
        super(Plot, self).calculate_right_margin()
        label_right = len(str(
            self.get_x_labels()[-1])) / 2 * self.font_size * 0.6
        self.border_right = max(label_right, self.border_right)

    def data_max(self, axis):
        data_index = getattr(self, '%s_data_index' % axis)
        max_value = max(chain(
            *map(lambda set: set['data'][data_index], self.data)))
        # above is same as
        #max_value = max(map(lambda set:
        #     max(set['data'][data_index]), self.data))
        spec_max = getattr(self, 'max_%s_value' % axis)
        # Python 3 doesn't allow comparing None to int, so use -∞
        if spec_max is None:
            spec_max = float('-Inf')
        max_value = max(max_value, spec_max)
        return max_value

    def data_min(self, axis):
        data_index = getattr(self, '%s_data_index' % axis)
        min_value = min(chain(
            *map(lambda set: set['data'][data_index], self.data)))
        spec_min = getattr(self, 'min_%s_value' % axis)
        if spec_min is not None:
            min_value = min(min_value, spec_min)
        return min_value

    x_data_index = 0
    y_data_index = 1

    def data_range(self, axis):
        side = {'x': 'right', 'y': 'top'}[axis]

        min_value = self.data_min(axis)
        max_value = self.data_max(axis)
        range = max_value - min_value

        side_pad = range / 20.0 or 10
        scale_range = (max_value + side_pad) - min_value

        scale_division = getattr(
            self, 'scale_%s_divisions' % axis) or (scale_range / 10.0)

        if getattr(self, 'scale_%s_integers' % axis):
            scale_division = round(scale_division) or 1

        return min_value, max_value, scale_division

    def x_range(self):
        return self.data_range('x')

    def y_range(self):
        return self.data_range('y')

    def get_data_values(self, axis):
        min_value, max_value, scale_division = self.data_range(axis)
        return tuple(float_range(*self.data_range(axis)))

    def get_x_values(self):
        return self.get_data_values('x')

    def get_y_values(self):
        return self.get_data_values('y')

    def get_x_labels(self):
        return map(str, self.get_x_values())

    def get_y_labels(self):
        return map(str, self.get_y_values())

    def field_size(self, axis):
        size = {'x': 'width', 'y': 'height'}[axis]
        side = {'x': 'right', 'y': 'top'}[axis]
        values = getattr(self, 'get_%s_values' % axis)()
        max_d = self.data_max(axis)
        dx = (
            float(max_d - values[-1]) / (values[-1] - values[-2])
            if len(values) > 1 else max_d
            )
        graph_size = getattr(self, 'graph_%s' % size)
        side_font = getattr(self, '%s_font' % side)
        side_align = getattr(self, '%s_align' % side)
        result = ((float(graph_size) - self.font_size * 2 * side_font) /
                  (len(values) + dx - side_align))
        return result

    def field_width(self):
        return self.field_size('x')

    def field_height(self):
        return self.field_size('y')

    def draw_data(self):
        self.load_transform_parameters()
        for line, data in izip(count(1), self.data):
            x_start, y_start = self.transform_output_coordinates(
                (data['data'][self.x_data_index][0],
                data['data'][self.y_data_index][0])
            )
            data_points = zip(*data['data'])
            graph_points = self.get_graph_points(data_points)
            lpath = self.get_lpath(graph_points)
            if self.area_fill:
                graph_height = self.graph_height
                path = etree.SubElement(self.graph, 'path', {
                    'd': 'M%(x_start)f %(graph_height)f'
                    ' %(lpath)s V%(graph_height)f Z' % vars(),
                    'class': 'fill%(line)d' % vars()})
            if self.draw_lines_between_points:
                path = etree.SubElement(self.graph, 'path', {
                    'd': 'M%(x_start)f %(y_start)f %(lpath)s' % vars(),
                    'class': 'line%(line)d' % vars()})
            self.draw_data_points(line, data_points, graph_points)
        self._draw_constant_lines()
        del self.__transform_parameters

    def add_constant_line(self, value, label=None, style=None):
        self.constant_lines = getattr(self, 'constant_lines', [])
        self.constant_lines.append((value, label, style))

    def _draw_constant_lines(self):
        if hasattr(self, 'constant_lines'):
            map(self.__draw_constant_line, self.constant_lines)

    def __draw_constant_line(self, value_label_style):
        "Draw a constant line on the y-axis with the label"
        value, label, style = value_label_style
        start = self.transform_output_coordinates((0, value))[1]
        stop = self.graph_width
        path = etree.SubElement(self.graph, 'path', {
            'd': 'M 0 %(start)s h%(stop)s' % vars(),
            'class': 'constantLine'})
        if style:
            path.set('style', style)
        text = etree.SubElement(self.graph, 'text', {
            'x': str(2),
            'y': str(start - 2),
            'class': 'constantLine'})
        text.text = label

    def load_transform_parameters(self):
        "Cache the parameters necessary to transform x & y coordinates"
        x_min, x_max, x_div = self.x_range()
        y_min, y_max, y_div = self.y_range()
        x_step = ((float(self.graph_width) - self.font_size * 2) /
                  (x_max - x_min))
        y_step = ((float(self.graph_height) - self.font_size * 2) /
                  (y_max - y_min))
        self.__transform_parameters = dict(vars())
        del self.__transform_parameters['self']

    def get_graph_points(self, data_points):
        return map(self.transform_output_coordinates, data_points)

    def get_lpath(self, points):
        points = map(lambda p: "%f %f" % p, points)
        return 'L' + ' '.join(points)

    def transform_output_coordinates(self, (x, y)):
        x_min = self.__transform_parameters['x_min']
        x_step = self.__transform_parameters['x_step']
        y_min = self.__transform_parameters['y_min']
        y_step = self.__transform_parameters['y_step']
        #locals().update(self.__transform_parameters)
        #vars().update(self.__transform_parameters)
        x = (x - x_min) * x_step
        y = self.graph_height - (y - y_min) * y_step
        return x, y

    def draw_data_points(self, line, data_points, graph_points):
        if not self.show_data_points and not self.show_data_values:
            return

        for ((dx, dy), (gx, gy)) in izip(data_points, graph_points):
            if self.show_data_points:
                etree.SubElement(self.graph, 'circle', {
                    'cx': str(gx),
                    'cy': str(gy),
                    'r': '2.5',
                    'class': 'dataPoint%(line)s' % vars()})
            if self.show_data_values:
                self.add_popup(gx, gy, self.format(dx, dy))
            self.make_datapoint_text(gx, gy - 6, dy)

    def format(self, x, y):
        return '(%0.2f, %0.2f)' % (x, y)