%matplotlib inlineimport matplotlib.pyplot as plt
import numpy as np
import pandas as pd
import seaborn as sns
np.random.seed(0)
plt.style.use('seaborn-whitegrid')
plt.rc('figure', figsize=(8, 5), facecolor='white')data = np.arange(10)
dataarray([0, 1, 2, 3, 4, 5, 6, 7, 8, 9])
plt.plot(data)[<matplotlib.lines.Line2D at 0x1a385c7610>]
Plots reside in a Figure object.
One or more subplots must be added.
Below, a 2x2 grid is created and the first subplot is set to ax1, the second to ax2, etc.
fig = plt.figure()
ax1 = fig.add_subplot(2, 2, 1)
ax2 = fig.add_subplot(2, 2, 2)
ax3 = fig.add_subplot(2, 2, 3)
When a plotting command is issued, such as plt.plot([1.5, 3.5, -2, 1.6]), the last figure and subplot used is drawn to.
fig = plt.figure()
ax1 = fig.add_subplot(2, 2, 1)
ax2 = fig.add_subplot(2, 2, 2)
plt.plot([1.5, 3.5, -2, 1.6])
ax3 = fig.add_subplot(2, 2, 3)
plt.plot(np.random.randn(50).cumsum(), 'k--')[<matplotlib.lines.Line2D at 0x1a389bf610>]
Alternatively, the objects returned by fig.add_subplot() can be added to directly.
fig = plt.figure()
ax1 = fig.add_subplot(2, 2, 1)
ax1.hist(np.random.randn(100), bins=20, color='k', alpha=0.3)
ax2 = fig.add_subplot(2, 2, 2)
plt.plot([1.5, 3.5, -2, 1.6])
ax3 = fig.add_subplot(2, 2, 3)
plt.plot(np.random.randn(50).cumsum(), 'k--')[<matplotlib.lines.Line2D at 0x1a38bb8f90>]
The author mentions that there are options in fig.subplot to declare that some subplots should have the same axes.
An example is shown in the next section.
The trick is to work on the axes array that is returned by plt.subplots().
fig, axes = plt.subplots(2, 3)
axesarray([[<matplotlib.axes._subplots.AxesSubplot object at 0x1a38a17d10>,
<matplotlib.axes._subplots.AxesSubplot object at 0x1a38cff790>,
<matplotlib.axes._subplots.AxesSubplot object at 0x1a38d30450>],
[<matplotlib.axes._subplots.AxesSubplot object at 0x1a38d6e7d0>,
<matplotlib.axes._subplots.AxesSubplot object at 0x1a38da4490>,
<matplotlib.axes._subplots.AxesSubplot object at 0x1a38de1810>]],
dtype=object)
All of the spacing is specified relative to the height and width of the plot.
fig, axes = plt.subplots(2, 2, sharex=True, sharey=True)
for i in range(2):
for j in range(2):
axes[i, j].hist(np.random.randn(500), bins=50, color='k', alpha=0.5)
plt.subplots_adjust(wspace=0, hspace=0)
The color and linestyle can be specified together or separately:
ax.plot(x, y, 'g--')
ax.plot(x, y, linestyle='--', color='g')Any color can be specified by its hex code.
Markers can also be used to highlight the data points.
plt.plot(np.random.randn(30).cumsum(), 'ko--')[<matplotlib.lines.Line2D at 0x1a39358c90>]
Which is equivalent to the more verbose statement below.
plt.plot(np.random.randn(30).cumsum(), color='k', linestyle='dashed', marker='o')[<matplotlib.lines.Line2D at 0x135df4e90>]
The type of interpolation between points on a line plot can be specified; the default is 'linear'.
data = np.random.randn(30).cumsum()
plt.plot(data, 'm--', label='Default')
plt.plot(data, 'b-', drawstyle='steps-post', label='steps-post')
plt.legend(loc='best')<matplotlib.legend.Legend at 0x135e5d390>
Ticks, labels, and legends
The general interactive API used for pyplot uses methods like xlim(), xticks(), and xticklabels() in two ways:
- called with no paramters, the current values are returned,
- called with paramteres, the values are set.
All such methods act on the active or most recently created AxesSubplot object.
These two methods actaully refer to two different methods on the subplot object, such as ax.get_xlim() and ax.set_xlim() - it is often recommended to explicitly use these.
fig = plt.figure()
ax = fig.add_subplot(1, 1, 1)
ax.plot(np.random.randn(1000).cumsum())
ticks = ax.set_xticks([0, 250, 500, 750, 1000])
labels = ax.set_xticklabels(['one', 'two', 'three', 'four', 'five'],
rotation=30,
fontsize='small')
ax.set_title('My fiirst matplotlib plot title')
ax.set_xlabel('Stages')Text(0.5, 0, 'Stages')
The easiest way of creating a legend is to pass a label argument to each plot.
def make_random_data(n=1000):
return np.random.randn(n).cumsum()
fig = plt.figure()
ax = fig.add_subplot(1, 1, 1)
ax.plot(make_random_data(), 'orange', label='one')
ax.plot(make_random_data(), 'magenta', label='two')
ax.plot(make_random_data(), 'purple', label='three')
ax.legend(loc='best')<matplotlib.legend.Legend at 0x1a395c9990>
Annotations can inclde text, arrows, or other shapes.
Some of the more common functions for such annotations include text(), arrow(), and annotate().
For an example, the author plots the S&P 500 closing prices since 2007 and adds some annotations for important dates of the 2008 financial crisis.
from datetime import datetime
data = pd.read_csv('assets/examples/spx.csv', index_col=0, parse_dates=True)
spx = data['SPX']
spx.head()1990-02-01 328.79
1990-02-02 330.92
1990-02-05 331.85
1990-02-06 329.66
1990-02-07 333.75
Name: SPX, dtype: float64
crisis_data = [
(datetime(2007, 10, 11), 'Peak of bull market'),
(datetime(2008, 3, 12), 'Bear Stearns Fails'),
(datetime(2008, 9, 15), 'Lehman Bankruptcy')
]
crisis_data[(datetime.datetime(2007, 10, 11, 0, 0), 'Peak of bull market'),
(datetime.datetime(2008, 3, 12, 0, 0), 'Bear Stearns Fails'),
(datetime.datetime(2008, 9, 15, 0, 0), 'Lehman Bankruptcy')]
fig = plt.figure()
ax = fig.add_subplot(1, 1, 1)
spx.plot(ax=ax, style='k-')
arrow_styling = {
'facecolor': 'black',
'headwidth': 4,
'width': 2,
'headlength': 4
}
for date, label in crisis_data:
ax.annotate(label, xy=(date, spx.asof(date) + 50),
xytext=(date, spx.asof(date) + 200),
arrowprops=arrow_styling,
horizontalalignment='left', verticalalignment='top')
ax.set_xlim(['1/1/2007', '1/1/2011'])
ax.set_ylim([600, 1800])
ax.set_title('Important dates in the 2008-2009 financial crisis')Text(0.5, 1.0, 'Important dates in the 2008-2009 financial crisis')
Use the plt.savefig('filepath.svg') method.
fig = plt.figure()
ax = fig.add_subplot(1, 1, 1)
spx.plot(ax=ax, style='k-')
arrow_styling = {
'facecolor': 'black',
'headwidth': 4,
'width': 2,
'headlength': 4
}
for date, label in crisis_data:
ax.annotate(label, xy=(date, spx.asof(date) + 50),
xytext=(date, spx.asof(date) + 200),
arrowprops=arrow_styling,
horizontalalignment='left', verticalalignment='top')
ax.set_xlim(['1/1/2007', '1/1/2011'])
ax.set_ylim([600, 1800])
ax.set_title('Important dates in the 2008-2009 financial crisis')
plt.savefig('assets/output/financial_crisis_annotated_plot.svg', facecolor='white')
Most default behavior can be customized via an extensive set of global parameters. The default plot size can be set like so.
The first argument is the component to be customized, followed by a sequence of keyward arguments indicating the new parameter values.
font_options = {
'family': 'monospace',
'weight': 'bold',
'size': 'small'
}
plt.rc('font', **font_options)Seaborn is a statistical graphics library aimed at simplifying many common visualization types.
Both Series and DataFrame have plot attributes to create some basic plot types.
By default, plot() creates a line plot.
s = pd.Series(np.random.randn(10).cumsum(), index=np.arange(0, 100, 10))
s.plot()<matplotlib.axes._subplots.AxesSubplot at 0x1a38653290>
Most of the pandas plotting methods accept a matplotlib subplot object through an optional ax parameter.
The plot() method for DataFrame makes line plot out of each column.
df = pd.DataFrame(np.random.randn(10, 4).cumsum(0),
columns=['A', 'B', 'C', 'D'],
index=np.arange(0, 100, 10))
df.plot()<matplotlib.axes._subplots.AxesSubplot at 0x1a39b22050>
The plot.bar() and plot.barh() make vertical and horizontal bar plots.
fig, axes = plt.subplots(2, 1)
data = pd.Series(np.random.rand(16), index=list('abcdefghijklmnop'))
data.plot.bar(ax=axes[0], color='k', alpha=0.7)
data.plot.barh(ax=axes[1], color='k', alpha=0.7)<matplotlib.axes._subplots.AxesSubplot at 0x1a39d13350>
A DataFrame bar plot groups the values of rows together.
df = pd.DataFrame(np.random.rand(6, 4),
index=['one', 'two', 'three', 'four', 'five', 'six'],
columns=pd.Index(['A', 'B', 'C', 'D'], name='Genus'))
df.dataframe tbody tr th {
vertical-align: top;
}
.dataframe thead th {
text-align: right;
}
| Genus | A | B | C | D |
|---|---|---|---|---|
| one | 0.424886 | 0.224244 | 0.762922 | 0.744435 |
| two | 0.647056 | 0.455640 | 0.464806 | 0.298444 |
| three | 0.265725 | 0.848971 | 0.859374 | 0.015345 |
| four | 0.719606 | 0.329922 | 0.919347 | 0.242120 |
| five | 0.335403 | 0.747336 | 0.074099 | 0.708039 |
| six | 0.674679 | 0.304022 | 0.712769 | 0.902925 |
df.plot.bar()
plt.legend(loc='best')<matplotlib.legend.Legend at 0x1a39db68d0>
df.plot.barh(stacked=True, alpha=0.7)
plt.legend(loc='best')<matplotlib.legend.Legend at 0x1a3a239050>
The author uses the tipping data as an example of plotting.
tips = pd.read_csv('assets/examples/tips.csv')
tips.head().dataframe tbody tr th {
vertical-align: top;
}
.dataframe thead th {
text-align: right;
}
| total_bill | tip | smoker | day | time | size | |
|---|---|---|---|---|---|---|
| 0 | 16.99 | 1.01 | No | Sun | Dinner | 2 |
| 1 | 10.34 | 1.66 | No | Sun | Dinner | 3 |
| 2 | 21.01 | 3.50 | No | Sun | Dinner | 3 |
| 3 | 23.68 | 3.31 | No | Sun | Dinner | 2 |
| 4 | 24.59 | 3.61 | No | Sun | Dinner | 4 |
# Number of group-size per day.
party_counts = pd.crosstab(tips['day'], tips['size'])
party_counts.dataframe tbody tr th {
vertical-align: top;
}
.dataframe thead th {
text-align: right;
}
| size | 1 | 2 | 3 | 4 | 5 | 6 |
|---|---|---|---|---|---|---|
| day | ||||||
| Fri | 1 | 16 | 1 | 1 | 0 | 0 |
| Sat | 2 | 53 | 18 | 13 | 1 | 0 |
| Sun | 0 | 39 | 15 | 18 | 3 | 1 |
| Thur | 1 | 48 | 4 | 5 | 1 | 3 |
# Normalize to sum to 1.
party_pcts = party_counts.div(party_counts.sum(1), axis=0)
party_pcts.dataframe tbody tr th {
vertical-align: top;
}
.dataframe thead th {
text-align: right;
}
| size | 1 | 2 | 3 | 4 | 5 | 6 |
|---|---|---|---|---|---|---|
| day | ||||||
| Fri | 0.052632 | 0.842105 | 0.052632 | 0.052632 | 0.000000 | 0.000000 |
| Sat | 0.022989 | 0.609195 | 0.206897 | 0.149425 | 0.011494 | 0.000000 |
| Sun | 0.000000 | 0.513158 | 0.197368 | 0.236842 | 0.039474 | 0.013158 |
| Thur | 0.016129 | 0.774194 | 0.064516 | 0.080645 | 0.016129 | 0.048387 |
party_pcts.plot.bar()<matplotlib.axes._subplots.AxesSubplot at 0x1a3a45e990>
The seaborn library can be useful for making comparisons. The following is an example looking at the average percent of the bill used for the tip per day of the week.
tips['tip_pct'] = tips['tip'] / (tips['total_bill'] - tips['tip'])
tips.head().dataframe tbody tr th {
vertical-align: top;
}
.dataframe thead th {
text-align: right;
}
| total_bill | tip | smoker | day | time | size | tip_pct | |
|---|---|---|---|---|---|---|---|
| 0 | 16.99 | 1.01 | No | Sun | Dinner | 2 | 0.063204 |
| 1 | 10.34 | 1.66 | No | Sun | Dinner | 3 | 0.191244 |
| 2 | 21.01 | 3.50 | No | Sun | Dinner | 3 | 0.199886 |
| 3 | 23.68 | 3.31 | No | Sun | Dinner | 2 | 0.162494 |
| 4 | 24.59 | 3.61 | No | Sun | Dinner | 4 | 0.172069 |
sns.barplot(x='tip_pct', y='day', data=tips, orient='h')<matplotlib.axes._subplots.AxesSubplot at 0x1354dbc10>
sns.barplot(x='tip_pct', y='day', hue='time', data=tips, orient='h')<matplotlib.axes._subplots.AxesSubplot at 0x1353b34d0>
tips['tip_pct'].plot.hist(bins=50)<matplotlib.axes._subplots.AxesSubplot at 0x123d10390>
tips['tip_pct'].plot.density()<matplotlib.axes._subplots.AxesSubplot at 0x1a3a7ddc10>
comp1 = np.random.normal(0, 1, size=200)
comp2 = np.random.normal(10, 2, size=200)
values = pd.Series(np.concatenate([comp1, comp2]))
sns.distplot(values, bins=100, color='k')<matplotlib.axes._subplots.AxesSubplot at 0x1a3c033050>
macro = pd.read_csv('assets/examples/macrodata.csv')
data = macro[['cpi', 'm1', 'tbilrate', 'unemp']]
data.head().dataframe tbody tr th {
vertical-align: top;
}
.dataframe thead th {
text-align: right;
}
| cpi | m1 | tbilrate | unemp | |
|---|---|---|---|---|
| 0 | 28.98 | 139.7 | 2.82 | 5.8 |
| 1 | 29.15 | 141.7 | 3.08 | 5.1 |
| 2 | 29.35 | 140.5 | 3.82 | 5.3 |
| 3 | 29.37 | 140.0 | 4.33 | 5.6 |
| 4 | 29.54 | 139.6 | 3.50 | 5.2 |
trans_data = np.log(data).diff().dropna()
trans_data.head().dataframe tbody tr th {
vertical-align: top;
}
.dataframe thead th {
text-align: right;
}
| cpi | m1 | tbilrate | unemp | |
|---|---|---|---|---|
| 1 | 0.005849 | 0.014215 | 0.088193 | -0.128617 |
| 2 | 0.006838 | -0.008505 | 0.215321 | 0.038466 |
| 3 | 0.000681 | -0.003565 | 0.125317 | 0.055060 |
| 4 | 0.005772 | -0.002861 | -0.212805 | -0.074108 |
| 5 | 0.000338 | 0.004289 | -0.266946 | 0.000000 |
sns.regplot('m1', 'unemp', data=trans_data)
plt.title(f'Changes in log m1 versus log unemp')Text(0.5, 1.0, 'Changes in log m1 versus log unemp')
sns.pairplot(trans_data, diag_kind='kde', plot_kws={'alpha': 0.2})<seaborn.axisgrid.PairGrid at 0x1a3d8d91d0>
The seaborn library has factorplot() for faceting plots.
sns.catplot(x='day', y='tip_pct',
row='time', col='smoker',
kind='bar',
data=tips[tips.tip_pct < 1])<seaborn.axisgrid.FacetGrid at 0x1a3e2b0f10>
sns.catplot(x='tip_pct', y='day',
kind='box',
data=tips[tips.tip_pct < 0.5])<seaborn.axisgrid.FacetGrid at 0x1a3d8b8250>
