Introductory Notes: Matplotlib

Preliminaries
Start by importing these Python modules
import numpy as np
import pandas as pd
from pandas import DataFrame, Series
import matplotlib.pyplot as plt
import matplotlib

Which Application Programming Interface?

The two worlds of Matplotlib
There are 2 broad ways of using pyplot:
1. The first (and most common) way is not pythonic. It
relies on global functions to build and display a global
figure using matplotlib as a global state machine.
(This is an easy approach for interactive use).
2. The second way is pythonic and object oriented. You
obtain an empty Figure from a global factory, and
then build the plot explicitly using the methods of the
Figure and the classes it contains. (This is the best
approach for programmatic use).
While these notes focus on second approach, let's begin
with a quick look at the first.

Using matplotlib in a non-pythonic way

1. Get some (fake) data - monthly time series
x = pd.period_range('1980-01-01',
periods=410, freq='M')
x = x.to_timestamp().to_pydatetime()
y = np.random.randn(len(x)).cumsum()
2. Plot the data
plt.plot(x, y, label='FDI')
3. Add your labels and pretty-up the plot
plt.title('Fake Data Index')
plt.xlabel('Date')
plt.ylabel('Index')
plt.grid(True)
plt.figtext(0.995, 0.01, 'Footnote',
ha='right', va='bottom')
plt.legend(loc='best', framealpha=0.5,
prop={'size':'small'})
plt.tight_layout(pad=1)
plt.gcf().set_size_inches(8, 4)
4. SAVE the figure
plt.savefig('filename.png')

Matplotlib: intro to the object oriented way

The Figure
Figure is the top-level container for everything on a
canvas. It was obtained from the global Figure factory.
fig = plt.figure(num=None, figsize=None,
dpi=None, facecolor=None,
edgecolor=None)
num integer or string identifier of figure
if num exists, it is selected
if num is None, a new one is allocated
figsize tuple of (width, height) in inches
dpi dots per inch
facecolor background; edgecolor border
Iterating over the open figures
for i in plt.get_fignums():
fig = plt.figure(i) # get the figure
print (fig.number) # do something
Close a figure
plt.close(fig.number) #
plt.close()
# close
plt.close(i)
# close
plt.close(name) # close
plt.close('all')# close

close figure
the current figure
figure numbered i
figure by str name
all figures

An Axes or Subplot (a subclass of Axes)

An Axes is a container class for a specific plot. A figure
may contain many Axes and/or Subplots. Subplots are
laid out in a grid within the Figure. Axes can be placed
anywhere on the Figure. There are a number of
methods that yield an Axes, including:
ax = fig.add_subplot(2,2,1) # row-col-num
ax = fig.add_axes([0.1,0.1,0.8,0.8])
All at once
We can use the subplots factory to get the Figure and all
the desired Axes at once.
fig, ax = plt.subplots()
fig,(ax1,ax2,ax3) = plt.subplots(nrows=3,
ncols=1, sharex=True, figsize=(8,4))

5. Finally, close the figure

plt.close()

Iterating the Axes within a Figure

for ax in fig.get_axes():
pass # do something

Alternatively, SHOW the figure

With IPython, follow steps 1 to 3 above then
plt.show() # Note: also closes the figure

Remove an Axes from a Figure

fig.delaxes(ax)

Version 3 May 2015 - [Draft Mark Graph mark dot the dot graph at gmail dot com @Mark_Graph on twitter]

1!

Line plots using ax.plot()

Single plot constructed with Figure and Axes
# --- get the data
x = np.linspace(0, 16, 800)
y = np.sin(x)
# --- get an empty figure and add an Axes
fig = plt.figure(figsize=(8,4))
ax = fig.add_subplot(1,1,1) # row-col-num
# --- line plot data on the Axes
ax.plot(x, y, 'b-', linewidth=2,
label=r'$y=\sin(x)$')
# --- add title, labels and legend, etc.
ax.set_ylabel(r'$y$', fontsize=16);
ax.set_xlabel(r'$x$', fontsize=16)
ax.legend(loc='best')
ax.grid(True)
fig.suptitle('The Sine Wave')
fig.tight_layout(pad=1)
fig.savefig('filename.png', dpi=125)

Multiple lines with markers on a line plot

# --- get the Figure and Axes all at once
fig, ax = plt.subplots(figsize=(8,4))
# --- plot some lines
N = 8 # the number of lines we will plot
styles = ['-', '--', '-.', ':']
markers = list('+ox^psDv')
x = np.linspace(0, 100, 20)
for i in range(N): # add line-by-line
y = x + x/5*i + i
s = styles[i % len(styles)]
m = markers[i % len(markers)]
ax.plot(x, y,
label='Line '+str(i+1)+' '+s+m,
marker=m, linewidth=2, linestyle=s)
# --- add grid, legend, title and save
ax.grid(True)
ax.legend(loc='best', prop={'size':'large'})
fig.suptitle('A Simple Line Plot')
fig.savefig('filename.png', dpi=125)

Scatter plots using ax.scatter()

A simple scatter plot
x = np.random.randn(100)
y = x + np.random.randn(100) + 10
fig, ax = plt.subplots(figsize=(8, 3))
ax.scatter(x, y, alpha=0.5, color='orchid')
fig.suptitle('Example Scatter Plot')
fig.tight_layout(pad=2);
ax.grid(True)
fig.savefig('filename1.png', dpi=125)

Add a regression line (using statsmodels)

import statsmodels.api as sm
x = sm.add_constant(x) # intercept
# Model: y ~ x + c
model = sm.OLS(y, x)
fitted = model.fit()
x_pred = np.linspace(x.min(), x.max(), 50)
x_pred2 = sm.add_constant(x_pred)
y_pred = fitted.predict(x_pred2)
ax.plot(x_pred, y_pred, '-',
color='darkorchid', linewidth=2)
fig.savefig('filename2.png', dpi=125)

Version 3 May 2015 - [Draft Mark Graph mark dot the dot graph at gmail dot com @Mark_Graph on twitter]

2!

Add confidence bands for the regression line

y_hat = fitted.predict(x)
y_err = y - y_hat
mean_x = x.T[1].mean()
n = len(x)
dof = n - fitted.df_model - 1
from scipy import stats
t = stats.t.ppf(1-0.025, df=dof) # 2-tail
s_err = np.sum(np.power(y_err, 2))
conf = t * np.sqrt((s_err/(n-2))*(1.0/n +
(np.power((x_pred-mean_x),2) /
((np.sum(np.power(x_pred,2))) n*(np.power(mean_x,2))))))
upper = y_pred + abs(conf)
lower = y_pred - abs(conf)
ax.fill_between(x_pred, lower, upper,
color='#888888', alpha=0.3)
fig.savefig('filename3.png', dpi=125)

Add a prediction interval for the regression line

from statsmodels.sandbox.regression.predstd\
import wls_prediction_std
sdev, lower, upper =
wls_prediction_std(fitted,
exog=x_pred2, alpha=0.05)
ax.fill_between(x_pred, lower, upper,
color='#888888', alpha=0.1)
fig.savefig('filename4.png', dpi=125)

Changing the marker size and colour

N = 100
x = np.random.rand(N)
y = np.random.rand(N)
size = ((np.random.rand(N) + 1) * 8) ** 2
colours = np.random.rand(N)
fig, ax = plt.subplots(figsize=(8,4))
l = ax.scatter(x, y, s=size, c=colours)
fig.colorbar(l)
ax.set_xlim((-0.05, 1.05))
ax.set_ylim((-0.05, 1.05))
fig.suptitle('Dramatic Scatter Plot')
fig.tight_layout(pad=2);
ax.grid(True)
fig.savefig('filename.png', dpi=125)
Note: matplotlib has a huge range of colour maps in
addition to the default used here.

Changing the marker symbol

fig, ax = plt.subplots(figsize=(8,5))
markers = list('ov^<>12348sphHdD+x*|_')
N = 10
for i, m in enumerate(markers):
x = np.arange(N)
y = np.repeat(i+1, N)
ax.scatter(x, y, marker=m, label=m,
s=50, c='cornflowerblue')
ax.set_xlim((-1,N))
ax.set_ylim((0,len(markers)+1))
ax.legend(loc='upper left', ncol=3,
prop={'size':'xx-large'},
shadow=True, title='Marker Legend')
ax.get_legend().get_title().set_color("red")
fig.suptitle('Markers ' +
'(with an oversized legend)')
fig.tight_layout(pad=2);
fig.savefig('filename.png', dpi=125)

Note: The confidence interval relates to the location of

the regression line. The predication interval relates to
the location of data points around the regression line.

Version 3 May 2015 - [Draft Mark Graph mark dot the dot graph at gmail dot com @Mark_Graph on twitter]

3!

Bar plots using ax.bar() and ax.barh()

A simple bar chart
The bars in a bar-plot are placed to the right of the bar xaxis location by default. Centred labels require a little
jiggling with the bar and label positions.
# --- get the data
N = 5
labels = list('ABCDEFGHIJKLM'[0:N])
data = np.array(range(N)) +
np.random.rand(N)
# --- plot the data
fig, ax = plt.subplots(figsize=(8, 3.5))
width = 0.8;
tickLocations = np.arange(N)
rectLocations = tickLocations-(width/2.0)
ax.bar(rectLocations, data, width,
color='wheat',
edgecolor='#8B7E66', linewidth=4.0)
# --- pretty-up the plot
ax.set_xticks(ticks= tickLocations)
ax.set_xticklabels(labels)
ax.set_xlim(min(tickLocations)-0.6,
max(tickLocations)+0.6)
ax.set_yticks(range(N)[1:])
ax.set_ylim((0,N))
ax.yaxis.grid(True)
# --- title and save
fig.suptitle("Bar Plot with " +
"Oversized Edges")
fig.tight_layout(pad=2)
fig.savefig('filename.png', dpi=125)

Side by side bar chart

# --- get the data
before = np.array([10, 11, 9, 12])
after = np.array([11, 12, 8, 17])
labels=['Group '+x for x in list('ABCD')]
# --- the plot left then right
fig, ax = plt.subplots(figsize=(8, 3.5))
width = 0.4 # bar width
xlocs = np.arange(len(before))
ax.bar(xlocs-width, before, width,
color='wheat', label='Males')
ax.bar(xlocs, after, width,
color='#8B7E66', label='Females')
# --- labels, grids and title, then save
ax.set_xticks(ticks=range(len(before)))
ax.set_xticklabels(labels)
ax.yaxis.grid(True)
ax.legend(loc='best')
ax.set_ylabel('Mean Group Result')
fig.suptitle('Group Results by Gender')
fig.tight_layout(pad=1)
fig.savefig('filename.png', dpi=125)

Stacked bar
# --- get some data
alphas = np.array( [23, 44, 52, 32] )
betas = np.array( [38, 49, 32, 61] )
labels = ['Sydney', 'Melb', 'Canb', 'Bris']
# --- the plot
fig, ax = plt.subplots(figsize=(8, 3.5))
width = 0.8;
xlocations=np.array(range(len(alphas)+2))
adjlocs = xlocations[1:-1] - width/2.0
ax.bar(adjlocs, alphas, width,
label='alpha', color='tan')
ax.bar(adjlocs, betas, width,
label='beta', color='wheat',
bottom=alphas)
# --- pretty-up and save
ax.set_xticks(ticks=xlocations[1:-1])
ax.set_xticklabels(labels)
ax.yaxis.grid(True)
ax.legend(loc='best', prop={'size':'small'})
fig.suptitle("Stacked Nonsense")
fig.tight_layout(pad=2)
fig.savefig('filename.png', dpi=125)

Horizontal bar charts

Just as tick placement needs to be managed with
vertical bars; so with horizontal bars (which are above
the y-tick mark)
labels = ['Males', 'Females', 'Persons']
data = [6.3, 7.2, 6.8]
width = 0.8
yTickPos = np.arange(len(data))
yBarPos = yTickPos - (width/2.0)
fig, ax = plt.subplots(figsize=(8, 3.5))
ax.barh(yBarPos,data,width,color='wheat')
ax.set_yticks(ticks= yTickPos)
ax.set_yticklabels(labels)
ax.set_ylim((min(yTickPos)-0.6,
max(yTickPos)+0.6))
ax.xaxis.grid(True)
ax.set_ylabel('Gender');
ax.set_xlabel('Rate (Percent)')
fig.suptitle("Horizontal Nonsense")
fig.tight_layout(pad=2)
fig.savefig('filename.png', dpi=125)

Version 3 May 2015 - [Draft Mark Graph mark dot the dot graph at gmail dot com @Mark_Graph on twitter]

4!

Pie Chart using ax.pie()

Plot spines
As nice as pie
# --- get some data
data = np.array([5,3,4,6])
labels = ['bats', 'cats', 'gnats', 'rats']
explode = (0, 0.1, 0, 0) # explode cats
colrs=['khaki', 'goldenrod', 'tan', 'wheat']
# --- the plot
fig, ax = plt.subplots(figsize=(8, 3.5))
ax.pie(data, explode=explode,
labels=labels, autopct='%1.1f%%',
startangle=270, colors=colrs)
ax.axis('equal') # keep it a circle
# --- tidy-up and save
fig.suptitle("Delicious Pie Ingredients")
fig.savefig('filename.png', dpi=125)

Hiding the top and right spines

x = np.linspace(-np.pi, np.pi, 800)
y = np.sin(x)
fig, ax = plt.subplots(figsize=(8, 4))
ax.plot(x, y, label='Sine', color='red')
ax.set_axis_bgcolor('#e5e5e5')
ax.spines['right'].set_color('none')
ax.spines['top'].set_color('none')
ax.spines['left'].set_position(
('outward',10))
ax.spines['bottom'].set_position(
('outward',10))
ax.xaxis.set_ticks_position('bottom')
ax.yaxis.set_ticks_position('left')
# do the ax.grid() after setting ticks
ax.grid(b=True, which='both',
color='white', linestyle='-',
linewidth=1.5)
ax.set_axisbelow(True)
ax.legend(loc='best', frameon=False)
fig.savefig('filename.png', dpi=125)

Polar using ax.plot()

Polar coordinates
# --- theta
theta = np.linspace(-np.pi, np.pi, 800)
# --- get us a Figure
fig = plt.figure(figsize=(8,4))
# --- left hand plot
ax = fig.add_subplot(1,2,1, polar=True)
r = 3 + np.cos(5*theta)
ax.plot(theta, r)
ax.set_yticks([1,2,3,4])
# --- right hand plot
ax = fig.add_subplot(1,2,2, polar=True)
r = (np.sin(theta)) - (np.cos(10*theta))
ax.plot(theta, r, color='green')
ax.set_yticks([1,2])
# --- title, explanatory text and save
fig.suptitle('Polar Coordinates')
fig.text(x=0.24, y=0.05,
s=r'$r = 3 + \cos(5 \theta)$')
fig.text(x=0.64, y=0.05,
s=r'$r = \sin(\theta) - \cos(10' +
r'\theta)$')
fig.savefig('filename.png', dpi=125)

Spines in the middle

x = np.linspace(-np.pi, np.pi, 800)
y = np.sin(x)
fig, ax = plt.subplots(figsize=(8, 4))
ax.plot(x, y, label='Sine')
ax.spines['right'].set_color('none')
ax.spines['top'].set_color('none')
ax.xaxis.set_ticks_position('bottom')
ax.spines['bottom'].set_position((
'data',0))
ax.yaxis.set_ticks_position('left')
ax.spines['left'].set_position((
'data',0))
ax.grid(b=True, which='both',
color='#888888', linestyle='-',
linewidth=0.5)
fig.suptitle('Sine')
fig.savefig('filename.png', dpi=125)

Version 3 May 2015 - [Draft Mark Graph mark dot the dot graph at gmail dot com @Mark_Graph on twitter]

5!

Legend to the right of the plot

N = 5
x = np.arange(N)
fig, ax = plt.subplots(figsize=(8, 3))
for j in range(5):
ax.plot(x, x*(j+1),
label='Line '+str(j))

Legends

box = ax.get_position() # Shrink plot

ax.set_position([box.x0, box.y0,
box.width * 0.8, box.height])
ax.legend(bbox_to_anchor=(1, 0.5),
loc='center left')
# Put legend
fig.savefig('filename.png', dpi=125)

Legend within the plot

Use the 'loc' argument to place the legend
N = 5
x = np.arange(N)
fig, ax = plt.subplots(figsize=(8, 3))
for j in range(5):
ax.plot(x, x*(j+1),label='Line'+str(j))
ax.legend(loc='upper left')
fig.savefig('filename.png', dpi=125)

Legend slightly outside of the plot

N = 5
x = np.arange(N)
fig, ax = plt.subplots(figsize=(8, 3))
for j in range(5):
ax.plot(x, x*(j+1),
label='Line '+str(j))

Legend below the plot

N = 5
x = np.arange(N)
fig, ax = plt.subplots(figsize=(8, 3))
for j in range(5):
ax.plot(x, x*(j+1),
label='Line '+str(j))
box = ax.get_position()
ax.set_position([box.x0,
box.y0 + box.height * 0.15,
box.width, box.height * 0.85])
ax.legend(bbox_to_anchor=(0.5, -0.075),
loc='upper center', ncol=N)
fig.savefig('filename.png', dpi=125)

ax.legend(bbox_to_anchor=(1.1, 1.05))
fig.savefig('filename.png', dpi=125)

Version 3 May 2015 - [Draft Mark Graph mark dot the dot graph at gmail dot com @Mark_Graph on twitter]

6!

Multiple plots on a canvas

Using Axes to place a plot within a plot
fig = plt.figure(figsize=(8,4))
fig.text(x=0.01, y=0.01, s='Figure',
color='#888888', ha='left',
va='bottom', fontsize=20)
# --- Main Axes
ax = fig.add_axes([0.1,0.1,0.8,0.8])
ax.text(x=0.01, y=0.01, s='Main Axes',
color='red', ha='left', va='bottom',
fontsize=20)
ax.set_xticks([]); ax.set_yticks([])
# --- Insert Axes
ax= fig.add_axes([0.15,0.65,0.2,0.2])
ax.text(x=0.01, y=0.01, s='Insert Axes',
color='blue', ha='left', va='bottom',
fontsize=20)
ax.set_xticks([]); ax.set_yticks([])
fig.suptitle('An Axes within an Axes')
fig.savefig('filename.png', dpi=125)

Simple subplot grid layouts

fig = plt.figure(figsize=(8,4))
fig.text(x=0.01, y=0.01, s='Figure',
color='#888888', ha='left',
va='bottom', fontsize=20)
for i in range(4):
# fig.add_subplot(nrows, ncols, num)
ax = fig.add_subplot(2, 2, i+1)
ax.text(x=0.01, y=0.01,
s='Subplot 2 2 '+str(i+1),
color='red', ha='left',
va='bottom', fontsize=20)
ax.set_xticks([]); ax.set_yticks([])
ax.set_xticks([]); ax.set_yticks([])
fig.suptitle('Subplots')
fig.savefig('filename.png', dpi=125)

Using GridSpec layouts (like list slicing)

import matplotlib.gridspec as gs
gs = gs.GridSpec(3, 3) # nrows, ncols
fig = plt.figure(figsize=(8,4))
fig.text(x=0.01, y=0.01, s='Figure',
color='#888888', ha='left',
va='bottom', fontsize=20)
ax1 = fig.add_subplot(gs[0, :]) # row,col
ax1.text(x=0.2,y=0.2,s='0, :', color='b')
ax2 = fig.add_subplot(gs[1,:-1])
ax2.text(x=0.2,y=0.2,s='1, :-1', color='b')
ax3 = fig.add_subplot(gs[1:, -1])
ax3.text(x=0.2,y=0.2, s='1:, -1', color='b')
ax4 = fig.add_subplot(gs[-1,0])
ax4.text(x=0.2,y=0.2, s='-1, :0', color='b')
ax5 = fig.add_subplot(gs[-1,-2])
ax5.text(x=0.2,y=0.2, s='-1,:-2', color='b')
for a in fig.get_axes():
a.set_xticks([])
a.set_yticks([])
fig.suptitle('GridSpec Layout')
fig.savefig('filename.png', dpi=125)

Plotting defaults
Configuration files
Matplotlib uses configuration files to set the defaults. So
that you can edit it, the location of the configuration file
can be found as follows:
print (matplotlib.matplotlib_fname())
Configuration settings
The current configuration settings
print (matplotlib.rcParams)
Change the default settings
plt.rc('figure', figsize=(8,4), dpi=125,
facecolor='white', edgecolor='white')
plt.rc('axes', facecolor='#e5e5e5',
grid=True, linewidth=1.0,
axisbelow=True)
plt.rc('grid', color='white', linestyle='-',
linewidth=2.0, alpha=1.0)
plt.rc('xtick', direction='out')
plt.rc('ytick', direction='out')
plt.rc('legend', loc='best')

Version 3 May 2015 - [Draft Mark Graph mark dot the dot graph at gmail dot com @Mark_Graph on twitter]

7!

Cautionary notes
This cheat sheet was cobbled together by bots roaming
the dark recesses of the Internet seeking ursine and
pythonic myths. There is no guarantee the narratives
were captured and transcribed accurately. You use
these notes at your own risk. You have been warned.

Version 3 May 2015 - [Draft Mark Graph mark dot the dot graph at gmail dot com @Mark_Graph on twitter]

8!