Matplotlib is hiring a Research Software Engineering Fellow! See discourse for details. Apply by January 3, 2020
有多种形状和尺寸的柱状图,带有matplotlib。
条形图对于可视化计数或带有误差条的汇总统计信息很有用。这些例子展示了使用matplotlib实现这一点的几种方法。
# Credit: Josh Hemann
import numpy as np
import matplotlib.pyplot as plt
from matplotlib.ticker import MaxNLocator
from collections import namedtuple
n_groups = 5
means_men = (20, 35, 30, 35, 27)
std_men = (2, 3, 4, 1, 2)
means_women = (25, 32, 34, 20, 25)
std_women = (3, 5, 2, 3, 3)
fig, ax = plt.subplots()
index = np.arange(n_groups)
bar_width = 0.35
opacity = 0.4
error_config = {'ecolor': '0.3'}
rects1 = ax.bar(index, means_men, bar_width,
alpha=opacity, color='b',
yerr=std_men, error_kw=error_config,
label='Men')
rects2 = ax.bar(index + bar_width, means_women, bar_width,
alpha=opacity, color='r',
yerr=std_women, error_kw=error_config,
label='Women')
ax.set_xlabel('Group')
ax.set_ylabel('Scores')
ax.set_title('Scores by group and gender')
ax.set_xticks(index + bar_width / 2)
ax.set_xticklabels(('A', 'B', 'C', 'D', 'E'))
ax.legend()
fig.tight_layout()
plt.show()
这个例子来自一个应用程序,在这个应用程序中,小学体育老师希望能够向家长展示他们的孩子在几次健身测试中的表现,重要的是,相对于其他孩子的表现。为了提取用于演示的绘图代码,我们只需要为小约翰尼·杜编一些数据…
Student = namedtuple('Student', ['name', 'grade', 'gender'])
Score = namedtuple('Score', ['score', 'percentile'])
# GLOBAL CONSTANTS
testNames = ['Pacer Test', 'Flexed Arm\n Hang', 'Mile Run', 'Agility',
'Push Ups']
testMeta = dict(zip(testNames, ['laps', 'sec', 'min:sec', 'sec', '']))
def attach_ordinal(num):
"""helper function to add ordinal string to integers
1 -> 1st
56 -> 56th
"""
suffixes = {str(i): v
for i, v in enumerate(['th', 'st', 'nd', 'rd', 'th',
'th', 'th', 'th', 'th', 'th'])}
v = str(num)
# special case early teens
if v in {'11', '12', '13'}:
return v + 'th'
return v + suffixes[v[-1]]
def format_score(scr, test):
"""
Build up the score labels for the right Y-axis by first
appending a carriage return to each string and then tacking on
the appropriate meta information (i.e., 'laps' vs 'seconds'). We
want the labels centered on the ticks, so if there is no meta
info (like for pushups) then don't add the carriage return to
the string
"""
md = testMeta[test]
if md:
return '{0}\n{1}'.format(scr, md)
else:
return scr
def format_ycursor(y):
y = int(y)
if y < 0 or y >= len(testNames):
return ''
else:
return testNames[y]
def plot_student_results(student, scores, cohort_size):
# create the figure
fig, ax1 = plt.subplots(figsize=(9, 7))
fig.subplots_adjust(left=0.115, right=0.88)
fig.canvas.set_window_title('Eldorado K-8 Fitness Chart')
pos = np.arange(len(testNames))
rects = ax1.barh(pos, [scores[k].percentile for k in testNames],
align='center',
height=0.5, color='m',
tick_label=testNames)
ax1.set_title(student.name)
ax1.set_xlim([0, 100])
ax1.xaxis.set_major_locator(MaxNLocator(11))
ax1.xaxis.grid(True, linestyle='--', which='major',
color='grey', alpha=.25)
# Plot a solid vertical gridline to highlight the median position
ax1.axvline(50, color='grey', alpha=0.25)
# set X-axis tick marks at the deciles
cohort_label = ax1.text(.5, -.07, 'Cohort Size: {0}'.format(cohort_size),
horizontalalignment='center', size='small',
transform=ax1.transAxes)
# Set the right-hand Y-axis ticks and labels
ax2 = ax1.twinx()
scoreLabels = [format_score(scores[k].score, k) for k in testNames]
# set the tick locations
ax2.set_yticks(pos)
# make sure that the limits are set equally on both yaxis so the
# ticks line up
ax2.set_ylim(ax1.get_ylim())
# set the tick labels
ax2.set_yticklabels(scoreLabels)
ax2.set_ylabel('Test Scores')
ax2.set_xlabel(('Percentile Ranking Across '
'{grade} Grade {gender}s').format(
grade=attach_ordinal(student.grade),
gender=student.gender.title()))
rect_labels = []
# Lastly, write in the ranking inside each bar to aid in interpretation
for rect in rects:
# Rectangle widths are already integer-valued but are floating
# type, so it helps to remove the trailing decimal point and 0 by
# converting width to int type
width = int(rect.get_width())
rankStr = attach_ordinal(width)
# The bars aren't wide enough to print the ranking inside
if width < 5:
# Shift the text to the right side of the right edge
xloc = width + 1
# Black against white background
clr = 'black'
align = 'left'
else:
# Shift the text to the left side of the right edge
xloc = 0.98*width
# White on magenta
clr = 'white'
align = 'right'
# Center the text vertically in the bar
yloc = rect.get_y() + rect.get_height()/2.0
label = ax1.text(xloc, yloc, rankStr, horizontalalignment=align,
verticalalignment='center', color=clr, weight='bold',
clip_on=True)
rect_labels.append(label)
# make the interactive mouse over give the bar title
ax2.fmt_ydata = format_ycursor
# return all of the artists created
return {'fig': fig,
'ax': ax1,
'ax_right': ax2,
'bars': rects,
'perc_labels': rect_labels,
'cohort_label': cohort_label}
student = Student('Johnny Doe', 2, 'boy')
scores = dict(zip(testNames,
(Score(v, p) for v, p in
zip(['7', '48', '12:52', '17', '14'],
np.round(np.random.uniform(0, 1,
len(testNames))*100, 0)))))
cohort_size = 62 # The number of other 2nd grade boys
arts = plot_student_results(student, scores, cohort_size)
plt.show()