colorline.py

# -*- coding: utf-8 -*-
# <nbformat>3.0</nbformat>

# <headingcell level=1>

# Examples of plotting (multi-)colored lines

# <markdowncell>

# Plotting multi-colored lines may be achieved using `LineCollection`.
# These examples use a more simple interface by introducing a function `colorline`.

# <codecell>

%pylab inline
# %config InlineBackend.figure_format = 'retina'  # Use for retina display resolution

# <codecell>

# Different styling for graphs (requires separate mpltools package)
# from mpltools import style
# style.use("ggplot")

# <headingcell level=2>

# Function definitions

# <codecell>

# Topics: line, color, LineCollection, cmap, colorline, codex
'''
Defines a function colorline that draws a (multi-)colored 2D line with coordinates x and y.
The color is taken from optional data in z, and creates a LineCollection.

z can be:
- empty, in which case a default coloring will be used based on the position along the input arrays
- a single number, for a uniform color [this can also be accomplished with the usual plt.plot]
- an array of the length of at least the same length as x, to color according to this data
- an array of a smaller length, in which case the colors are repeated along the curve

The function colorline returns the LineCollection created, which can be modified afterwards.

See also: plt.streamplot
'''

import numpy as np
import matplotlib.pyplot as plt
from matplotlib.collections import LineCollection
from matplotlib.colors import ListedColormap, BoundaryNorm


# Data manipulation:

def make_segments(x, y):
    '''
    Create list of line segments from x and y coordinates, in the correct format for LineCollection:
    an array of the form   numlines x (points per line) x 2 (x and y) array
    '''

    points = np.array([x, y]).T.reshape(-1, 1, 2)
    segments = np.concatenate([points[:-1], points[1:]], axis=1)
    
    return segments


# Interface to LineCollection:

def colorline(x, y, z=None, cmap=plt.get_cmap('copper'), norm=plt.Normalize(0.0, 1.0), linewidth=3, alpha=1.0):
    '''
    Plot a colored line with coordinates x and y
    Optionally specify colors in the array z
    Optionally specify a colormap, a norm function and a line width
    '''
    
    # Default colors equally spaced on [0,1]:
    if z is None:
        z = np.linspace(0.0, 1.0, len(x))
           
    # Special case if a single number:
    if not hasattr(z, "__iter__"):  # to check for numerical input -- this is a hack
        z = np.array([z])
        
    z = np.asarray(z)
    
    segments = make_segments(x, y)
    lc = LineCollection(segments, array=z, cmap=cmap, norm=norm, linewidth=linewidth, alpha=alpha)
    
    ax = plt.gca()
    ax.add_collection(lc)
    
    return lc
        
    
def clear_frame(ax=None): 
    # Taken from a post by Tony S Yu
    if ax is None: 
        ax = plt.gca() 
    ax.xaxis.set_visible(False) 
    ax.yaxis.set_visible(False) 
    for spine in ax.spines.itervalues(): 
        spine.set_visible(False) 

# <headingcell level=2>

# Example 1: Sine wave colored by time (uses the defaults for colorline)

# <codecell>

# Sine wave colored by time

x = np.linspace(0, 4.*np.pi, 1000)
y = np.sin(x)

fig, axes = plt.subplots()

colorline(x, y)

plt.xlim(x.min(), x.max())
plt.ylim(-1.0, 1.0)
plt.show()

# <headingcell level=2>

# Example 2: Different colors for shifted sine waves

# <codecell>

# Shifting a sine wave: Example modified from mpltools by Tony S Yu

x = np.linspace(0, 4.*np.pi, 1000)
y = np.sin(x)

fig, axes = plt.subplots()

N = 10
for i in xrange(N):
    color = i / float(N)
    shift = 0.2 * i
    colorline(x-shift, y, color, cmap="cool")
    
plt.xlim(x.min(), x.max())
plt.ylim(-1.0, 1.0)
plt.show()  

# <headingcell level=2>

# Example 3: A polar function, plotted with a custom colored dash style

# <codecell>

# Polar function:

max_t = 6*np.pi
theta = np.linspace(0, max_t, 1000)

alpha = 0.3
r = 1 + alpha*np.sin(7./3. * theta)

x = r * np.sin(theta)
y = r * np.cos(theta)

colored_dashes = [0.1]*8 + [0.5]*4 + [0.8]*2  # uses Python list syntax

fig, axes = plt.subplots(figsize=(6,6))

colorline(x, y, colored_dashes, cmap='jet', linewidth=8) #, [0.3]*8+[0.4]*4+[0.5]*2)

plt.xlim(x.min() - 0.1, x.max() + 0.1)
plt.ylim(y.min() - 0.1, y.max() + 0.1)

plt.axis('equal')
clear_frame()
plt.show()

# <headingcell level=2>

# Example 4: A non-periodic Lissajous curve, colored with a continuous palette and different line widths

# <codecell>

# Non-periodic Lissajous curve

max_t = 15.
t = np.linspace(0, max_t, 5000)
x = np.sin(t)
y = np.sin(t * np.pi)

fig, axes = plt.subplots(figsize=(6,6))

lc = colorline(x, y, alpha=0.75, linewidth=t, cmap='coolwarm') 
# We can capture the LineCollection created to modify it later, e.g.:
# lc.set_rasterized = False

plt.xlim(x.min() - 0.1, x.max() + 0.1)
plt.ylim(y.min() - 0.1, y.max() + 0.1)

# plt.colorbar()  # Does not work at the moment!
plt.show()

# <headingcell level=2>

# Example 5: $\sin(1/x)$, colored according to its derivative and plotted using logarithmic spacing

# <codecell>

# sin(1/x), colored according to the value of its derivative

# Use logarithmic spacing so that looks better:
log_x = np.linspace(np.log(0.005), np.log(0.1), 5000)
x = np.exp(log_x)

y = 0.9 * np.sin(1. / x)

# Use finite differences to calculate the derivative of y:
y_deriv = (y[1:] - y[:-1]) / (x[1:] - x[:-1]) 

# Create a colormap for red, green and blue:
cmap = ListedColormap(['r', 'g', 'b'])

# Create a 'norm' (normalizing function) which maps data values to the interval [0,1]:
norm = BoundaryNorm([-1000, -200, 200, 1000], cmap.N)  # cmap.N is number of items in the colormap

fig, axes = plt.subplots()

colorline(x, y, y_deriv, cmap=cmap, norm=norm, linewidth=100.*x)
# This works even though y_deriv is a different length -- wraps around

plt.xlim(x.min(), x.max())
plt.ylim(-1.0, 1.0)
plt.show()