Skip to content

Navigation Menu

Sign in
Appearance settings

Search code, repositories, users, issues, pull requests...

Provide feedback

We read every piece of feedback, and take your input very seriously.

Saved searches

Use saved searches to filter your results more quickly
Appearance settings

Commit 5aadd99

Browse filesBrowse files
phobsonphobson
committed
DOC: MEP12 the histogram examples
1 parent 26e762d commit 5aadd99
Copy full SHA for 5aadd99

File tree

Expand file treeCollapse file tree

5 files changed

+91
-35
lines changed
Filter options
Expand file treeCollapse file tree

5 files changed

+91
-35
lines changed

‎examples/statistics/histogram_demo_cumulative.py

Copy file name to clipboard
+41-9Lines changed: 41 additions & 9 deletions
Original file line numberDiff line numberDiff line change
@@ -1,29 +1,61 @@
1-
"""
2-
Demo of the histogram (hist) function used to plot a cumulative distribution.
1+
"""Demo of using histograms to plot a cumulative distribution.
2+
3+
This shows how to plot a cumulative, normalized histogram as a
4+
step function as means of visualization the empirical cumulative
5+
distribution function (CDF) of a sample. We also use the `mlab`
6+
module to show the theoretical CDF.
7+
8+
A couple of other options to the `hist` function are demostrated.
9+
Namely, we use the `normed` parameter to normalize the histogram and
10+
a couple of different options to the `cumulative` parameter. Normalizing
11+
a histogram means that the counts within each bin are scaled such that
12+
the total height of each bin sum up to 1. Since we're showing the
13+
cumulative histogram, the max value at the end of the series is 1.
14+
The `normed` parameter takes a boolean value.
15+
16+
The `cumulative` kwarg is a little more nuanced. Like `normed`, you can
17+
pass it True or False, but you can also pass it -1 and that will
18+
reverse the distribution. In engineering, CDFs where `cumulative` is
19+
simply True are sometimes "non-excedance" curves. In other words, you
20+
can look at the y-value to set the probability of excedance. For example
21+
the value of 225 on the x-axis corresponse to about 0.85 on the y-axis,
22+
so there's an 85% chance that an observation in the sames does not
23+
exceed 225.
24+
25+
Conversely, setting, `cumulative` to -1 as is done in the last series
26+
for this example, creates a "excedance" curve.
327
428
"""
529
import numpy as np
630
import matplotlib.pyplot as plt
731
from matplotlib import mlab
32+
np.random.seed(0)
833

934

1035
mu = 200
1136
sigma = 25
1237
n_bins = 50
13-
x = mu + sigma*np.random.randn(10000)
38+
x = mu + sigma*np.random.randn(100)
1439

15-
n, bins, patches = plt.hist(x, n_bins, normed=1,
16-
histtype='step', cumulative=True)
40+
n, bins, patches = ax.hist(x, n_bins, normed=1, histtype='step',
41+
cumulative=True, label='Empirical distribution')
1742

1843
# Add a line showing the expected distribution.
1944
y = mlab.normpdf(bins, mu, sigma).cumsum()
2045
y /= y[-1]
21-
plt.plot(bins, y, 'k--', linewidth=1.5)
46+
47+
fig, ax = plt.subplots()
48+
ax.plot(bins, y, 'k--', linewidth=1.5, labe='Theoretical distribution')
2249

2350
# Overlay a reversed cumulative histogram.
24-
plt.hist(x, bins=bins, normed=1, histtype='step', cumulative=-1)
51+
ax.hist(x, bins=bins, normed=1, histtype='step', cumulative=-1,
52+
label='Reversed empirical distribution')
53+
54+
ax.grid(True)
55+
ax.legend(loc='right')
56+
ax.set_title('Cumulative step histograms')
2557

26-
plt.grid(True)
27-
plt.title('cumulative step')
58+
ax.set_xlabel('Annual rainfall (mm)')
59+
ax.set_ylabel('Likelihood of occurance')
2860

2961
plt.show()

‎examples/statistics/histogram_demo_features.py

Copy file name to clipboard
+18-11Lines changed: 18 additions & 11 deletions
Original file line numberDiff line numberDiff line change
@@ -1,35 +1,42 @@
11
"""
22
Demo of the histogram (hist) function with a few features.
33
4-
In addition to the basic histogram, this demo shows a few optional features:
4+
In addition to the basic histogram, this demo shows a few optional
5+
features:
56
67
* Setting the number of data bins
7-
* The ``normed`` flag, which normalizes bin heights so that the integral of
8-
the histogram is 1. The resulting histogram is a probability density.
8+
* The ``normed`` flag, which normalizes bin heights so that the
9+
integral of the histogram is 1. The resulting histogram is an
10+
approximation of the probability density function.
911
* Setting the face color of the bars
1012
* Setting the opacity (alpha value).
11-
1213
"""
14+
1315
import numpy as np
1416
import matplotlib.mlab as mlab
1517
import matplotlib.pyplot as plt
18+
np.random.seed(0)
1619

1720

1821
# example data
1922
mu = 100 # mean of distribution
2023
sigma = 15 # standard deviation of distribution
21-
x = mu + sigma * np.random.randn(10000)
24+
x = mu + sigma * np.random.randn(10)
2225

2326
num_bins = 50
27+
28+
fig, ax = plt.subplots()
29+
2430
# the histogram of the data
25-
n, bins, patches = plt.hist(x, num_bins, normed=1)
31+
n, bins, patches = ax.hist(x, num_bins, normed=1)
32+
2633
# add a 'best fit' line
2734
y = mlab.normpdf(bins, mu, sigma)
28-
plt.plot(bins, y, '--')
29-
plt.xlabel('Smarts')
30-
plt.ylabel('Probability density')
31-
plt.title(r'Histogram of IQ: $\mu=100$, $\sigma=15$')
35+
ax.plot(bins, y, '--')
36+
ax.set_xlabel('Smarts')
37+
ax.set_ylabel('Probability density')
38+
ax.set_title(r'Histogram of IQ: $\mu=100$, $\sigma=15$')
3239

3340
# Tweak spacing to prevent clipping of ylabel
34-
plt.subplots_adjust(left=0.15)
41+
fig.tight_layout()
3542
plt.show()

‎examples/statistics/histogram_demo_histtypes.py

Copy file name to clipboard
+6-4Lines changed: 6 additions & 4 deletions
Original file line numberDiff line numberDiff line change
@@ -1,17 +1,19 @@
11
"""
2-
Demo of the histogram (hist) function with different ``histtype`` settings.
2+
Demo of the histogram (hist) function with different ``histtype``
3+
settings.
34
45
* Histogram with step curve that has a color fill.
5-
* Histogram with with unequal bin widths.
6+
* Histogram with custom and unequal bin widths.
67
78
"""
89
import numpy as np
910
import matplotlib.pyplot as plt
11+
np.random.seed(0)
1012

1113

1214
mu = 200
1315
sigma = 25
14-
x = mu + sigma*np.random.randn(10000)
16+
x = mu + sigma*np.random.randn(100)
1517

1618
fig, (ax0, ax1) = plt.subplots(ncols=2, figsize=(8, 4))
1719

@@ -23,5 +25,5 @@
2325
ax1.hist(x, bins, normed=1, histtype='bar', rwidth=0.8)
2426
ax1.set_title('unequal bins')
2527

26-
plt.tight_layout()
28+
fig.tight_layout()
2729
plt.show()

‎examples/statistics/histogram_demo_multihist.py

Copy file name to clipboardExpand all lines: examples/statistics/histogram_demo_multihist.py
+4-3Lines changed: 4 additions & 3 deletions
Original file line numberDiff line numberDiff line change
@@ -8,15 +8,16 @@
88
* Step curve with a color fill
99
* Data sets of different sample sizes
1010
"""
11+
1112
import numpy as np
1213
import matplotlib.pyplot as plt
13-
14+
np.random.seed(0)
1415

1516
n_bins = 10
1617
x = np.random.randn(1000, 3)
1718

1819
fig, axes = plt.subplots(nrows=2, ncols=2)
19-
ax0, ax1, ax2, ax3 = axes.flat
20+
ax0, ax1, ax2, ax3 = axes.flatten()
2021

2122
colors = ['red', 'tan', 'lime']
2223
ax0.hist(x, n_bins, normed=1, histtype='bar', color=colors, label=colors)
@@ -34,5 +35,5 @@
3435
ax3.hist(x_multi, n_bins, histtype='bar')
3536
ax3.set_title('different sample sizes')
3637

37-
plt.tight_layout()
38+
fig.tight_layout()
3839
plt.show()

‎examples/statistics/multiple_histograms_side_by_side.py

Copy file name to clipboardExpand all lines: examples/statistics/multiple_histograms_side_by_side.py
+22-8Lines changed: 22 additions & 8 deletions
Original file line numberDiff line numberDiff line change
@@ -1,36 +1,50 @@
11

2-
"""
3-
Demo of how to produce multiple histograms side by side
2+
"""Demo of how to produce multiple histograms side by side
3+
4+
This example plots horizonal histograms of different samples along
5+
a categorical x-axis. Additionally, the histograms are plotted to
6+
be symmetrical about their x-position, thus making them very similar
7+
to violin plots.
8+
9+
To make this highly specialized plot, we can't use the standard `hist`
10+
method. Instead we use `barh` to draw the horizontal bars directly. The
11+
vertical positions and lengths of the bars are computed via the
12+
`np.histogram` function. The histograms for all the samples are
13+
computed using the same range (min and max values) and number of bins,
14+
so that the bins for each sample are in the same vertical positions.
415
"""
516

617
import numpy as np
718
import matplotlib.pyplot as plt
19+
np.random.seed(0)
820

921
number_of_bins = 20
1022

1123
# An example of three data sets to compare
12-
number_of_data_points = 1000
24+
number_of_data_points = 87
1325
labels = ["A", "B", "C"]
1426
data_sets = [np.random.normal(0, 1, number_of_data_points),
1527
np.random.normal(6, 1, number_of_data_points),
1628
np.random.normal(-3, 1, number_of_data_points)]
1729

1830
# Computed quantities to aid plotting
1931
hist_range = (np.min(data_sets), np.max(data_sets))
20-
binned_data_sets = [np.histogram(d, range=hist_range, bins=number_of_bins)[0]
21-
for d in data_sets]
32+
binned_data_sets = [
33+
np.histogram(d, range=hist_range, bins=number_of_bins)[0]
34+
for d in data_sets
35+
]
2236
binned_maximums = np.max(binned_data_sets, axis=1)
2337
x_locations = np.arange(0, sum(binned_maximums), np.max(binned_maximums))
2438

2539
# The bin_edges are the same for all of the histograms
2640
bin_edges = np.linspace(hist_range[0], hist_range[1], number_of_bins + 1)
27-
centers = .5 * (bin_edges + np.roll(bin_edges, 1))[:-1]
41+
centers = 0.5 * (bin_edges + np.roll(bin_edges, 1))[:-1]
2842
heights = np.diff(bin_edges)
2943

3044
# Cycle through and plot each histogram
31-
ax = plt.subplot(111)
45+
fig, ax = plt.subplots()
3246
for x_loc, binned_data in zip(x_locations, binned_data_sets):
33-
lefts = x_loc - .5 * binned_data
47+
lefts = x_loc - 0.5 * binned_data
3448
ax.barh(centers, binned_data, height=heights, left=lefts)
3549

3650
ax.set_xticks(x_locations)

0 commit comments

Comments
0 (0)
Morty Proxy This is a proxified and sanitized view of the page, visit original site.