数据科学和人工智能技术笔记二十、数据可视化-摩杜云开发者社区

二十、数据可视化

作者：Chris Albon

译者：飞龙

协议：CC BY-NC-SA 4.0

MatPlotLib 中的双向条形图

%matplotlib inline
import pandas as pd
import matplotlib.pyplot as plt
import numpy as np

# 创建数据帧
raw_data = {'first_name': ['Jason', 'Molly', 'Tina', 'Jake', 'Amy'],
        'pre_score': [4, 24, 31, 2, 3],
        'mid_score': [25, 94, 57, 62, 70],
        'post_score': [5, 43, 23, 23, 51]}
df = pd.DataFrame(raw_data, columns = ['first_name', 'pre_score', 'mid_score', 'post_score'])
df

	first_name	pre_score	mid_score	post_score
0	Jason	4	25	5
1	Molly	24	94	43
2	Tina	31	57	23
3	Jake	2	62	23
4	Amy	3	70	51

# 输入数据，特别是第二和
# 第三行，跳过第一列
x1 = df.ix[1, 1:]
x2 = df.ix[2, 1:]

# 创建条形标签
bar_labels = ['Pre Score', 'Mid Score', 'Post Score']

# 创建图形
fig = plt.figure(figsize=(8,6))

# 设置 y 的位置
y_pos = np.arange(len(x1))
y_pos = [x for x in y_pos]
plt.yticks(y_pos, bar_labels, fontsize=10)

# 在 y_pos 的位置上创建水平条形
plt.barh(y_pos, 
         # 使用数据 x1
         x1, 
         # 中心对齐
         align='center', 
         # 透明度为 0.4
         alpha=0.4, 
         # 颜色为绿色
         color='#263F13')

# 在 y_pos 的位置上创建水平条形
plt.barh(y_pos, 
         # 使用数据 -x2
         -x2,
         # 中心对齐
         align='center', 
         # 透明度为 0.4
         alpha=0.4, 
         # 颜色为绿色
         color='#77A61D')

# 注解和标签
plt.xlabel('Tina\'s Score: Light Green. Molly\'s Score: Dark Green')
t = plt.title('Comparison of Molly and Tina\'s Score')
plt.ylim([-1,len(x1)+0.1])
plt.xlim([-max(x2)-10, max(x1)+10])
plt.grid()

plt.show()

MatPlotLib 中的条形图

%matplotlib inline
import pandas as pd
import matplotlib.pyplot as plt
import numpy as np

# 创建数据帧
raw_data = {'first_name': ['Jason', 'Molly', 'Tina', 'Jake', 'Amy'],
        'pre_score': [4, 24, 31, 2, 3],
        'mid_score': [25, 94, 57, 62, 70],
        'post_score': [5, 43, 23, 23, 51]}
df = pd.DataFrame(raw_data, columns = ['first_name', 'pre_score', 'mid_score', 'post_score'])
df

	first_name	pre_score	mid_score	post_score
0	Jason	4	25	5
1	Molly	24	94	43
2	Tina	31	57	23
3	Jake	2	62	23
4	Amy	3	70	51

# 为每个变量创建得分均值的列表
mean_values = [df['pre_score'].mean(), df['mid_score'].mean(), df['post_score'].mean()]

# 创建变动列表，设为得分上下 .25
variance = [df['pre_score'].mean() * 0.25, df['pre_score'].mean() * 0.25, df['pre_score'].mean() * 0.25]

# 设置条形标签
bar_labels = ['Pre Score', 'Mid Score', 'Post Score']

# 创建条形的 x 位置
x_pos = list(range(len(bar_labels)))

# 在 x 位置上创建条形图
plt.bar(x_pos,
        # 使用 mean_values 中的数据
        mean_values, 
        # y-error 直线设置为变动
        yerr=variance, 
        # 中心对齐
        align='center',
        # 颜色
        color='#FFC222',
        # 透明度为 0.5
        alpha=0.5)

# 添加网格
plt.grid()

# 设置 y 轴高度
max_y = max(zip(mean_values, variance)) # returns a tuple, here: (3, 5)
plt.ylim([0, (max_y[0] + max_y[1]) * 1.1])

# 设置轴标签和标题
plt.ylabel('Score')
plt.xticks(x_pos, bar_labels)
plt.title('Mean Scores For Each Test')

plt.show()

Seaborn 中的调色板

import pandas as pd
%matplotlib inline
import matplotlib.pyplot as plt
import seaborn as sns

# 创建数据帧
data = {'date': ['2014-05-01 18:47:05.069722', '2014-05-01 18:47:05.119994', '2014-05-02 18:47:05.178768', '2014-05-02 18:47:05.230071', '2014-05-02 18:47:05.230071', '2014-05-02 18:47:05.280592', '2014-05-03 18:47:05.332662', '2014-05-03 18:47:05.385109', '2014-05-04 18:47:05.436523', '2014-05-04 18:47:05.486877'], 
        'deaths_regiment_1': [34, 43, 14, 15, 15, 14, 31, 25, 62, 41],
        'deaths_regiment_2': [52, 66, 78, 15, 15, 5, 25, 25, 86, 1],
        'deaths_regiment_3': [13, 73, 82, 58, 52, 87, 26, 5, 56, 75],
        'deaths_regiment_4': [44, 75, 26, 15, 15, 14, 54, 25, 24, 72],
        'deaths_regiment_5': [25, 24, 25, 15, 57, 68, 21, 27, 62, 5],
        'deaths_regiment_6': [84, 84, 26, 15, 15, 14, 26, 25, 62, 24],
        'deaths_regiment_7': [46, 57, 26, 15, 15, 14, 26, 25, 62, 41]}
df = pd.DataFrame(data, columns = ['date', 'battle_deaths', 'deaths_regiment_1', 'deaths_regiment_2',
                                   'deaths_regiment_3', 'deaths_regiment_4', 'deaths_regiment_5',
                                   'deaths_regiment_6', 'deaths_regiment_7'])
df = df.set_index(df.date)

sns.palplot(sns.color_palette("deep", 10))

sns.palplot(sns.color_palette("muted", 10))

sns.palplot(sns.color_palette("bright", 10))

sns.palplot(sns.color_palette("dark", 10))

sns.palplot(sns.color_palette("colorblind", 10))

sns.palplot(sns.color_palette("Paired", 10))

sns.palplot(sns.color_palette("BuGn", 10))

sns.palplot(sns.color_palette("GnBu", 10))

sns.palplot(sns.color_palette("OrRd", 10))

sns.palplot(sns.color_palette("PuBu", 10))

sns.palplot(sns.color_palette("YlGn", 10))

sns.palplot(sns.color_palette("YlGnBu", 10))

sns.palplot(sns.color_palette("YlOrBr", 10))

sns.palplot(sns.color_palette("YlOrRd", 10))

sns.palplot(sns.color_palette("BrBG", 10))

sns.palplot(sns.color_palette("PiYG", 10))

sns.palplot(sns.color_palette("PRGn", 10))

sns.palplot(sns.color_palette("PuOr", 10))

sns.palplot(sns.color_palette("RdBu", 10))

sns.palplot(sns.color_palette("RdGy", 10))

sns.palplot(sns.color_palette("RdYlBu", 10))

sns.palplot(sns.color_palette("RdYlGn", 10))

sns.palplot(sns.color_palette("Spectral", 10))

# 创建调色板并将其设为当前调色板
flatui = ["#9b59b6", "#3498db", "#95a5a6", "#e74c3c", "#34495e", "#2ecc71"]
sns.set_palette(flatui)
sns.palplot(sns.color_palette())

# 设置绘图颜色
sns.tsplot([df.deaths_regiment_1, df.deaths_regiment_2, df.deaths_regiment_3, df.deaths_regiment_4,
            df.deaths_regiment_5, df.deaths_regiment_6, df.deaths_regiment_7], color="#34495e")

# <matplotlib.axes._subplots.AxesSubplot at 0x116f5db70>

使用 Seaborn 和 pandas 创建时间序列绘图

import pandas as pd
%matplotlib inline
import matplotlib.pyplot as plt
import seaborn as sns

data = {'date': ['2014-05-01 18:47:05.069722', '2014-05-01 18:47:05.119994', '2014-05-02 18:47:05.178768', '2014-05-02 18:47:05.230071', '2014-05-02 18:47:05.230071', '2014-05-02 18:47:05.280592', '2014-05-03 18:47:05.332662', '2014-05-03 18:47:05.385109', '2014-05-04 18:47:05.436523', '2014-05-04 18:47:05.486877'], 
        'deaths_regiment_1': [34, 43, 14, 15, 15, 14, 31, 25, 62, 41],
        'deaths_regiment_2': [52, 66, 78, 15, 15, 5, 25, 25, 86, 1],
        'deaths_regiment_3': [13, 73, 82, 58, 52, 87, 26, 5, 56, 75],
        'deaths_regiment_4': [44, 75, 26, 15, 15, 14, 54, 25, 24, 72],
        'deaths_regiment_5': [25, 24, 25, 15, 57, 68, 21, 27, 62, 5],
        'deaths_regiment_6': [84, 84, 26, 15, 15, 14, 26, 25, 62, 24],
        'deaths_regiment_7': [46, 57, 26, 15, 15, 14, 26, 25, 62, 41]}
df = pd.DataFrame(data, columns = ['date', 'battle_deaths', 'deaths_regiment_1', 'deaths_regiment_2',
                                   'deaths_regiment_3', 'deaths_regiment_4', 'deaths_regiment_5',
                                   'deaths_regiment_6', 'deaths_regiment_7'])
df = df.set_index(df.date)

sns.tsplot([df.deaths_regiment_1, df.deaths_regiment_2, df.deaths_regiment_3, df.deaths_regiment_4,
            df.deaths_regiment_5, df.deaths_regiment_6, df.deaths_regiment_7], color="indianred")

# <matplotlib.axes._subplots.AxesSubplot at 0x1140be780>

# 带有置信区间直线，但是没有直线的时间序列绘图
sns.tsplot([df.deaths_regiment_1, df.deaths_regiment_2, df.deaths_regiment_3, df.deaths_regiment_4,
            df.deaths_regiment_5, df.deaths_regiment_6, df.deaths_regiment_7], err_style="ci_bars", interpolate=False)

# <matplotlib.axes._subplots.AxesSubplot at 0x116400668>

使用 Seaborn 创建散点图

import pandas as pd
%matplotlib inline
import random
import matplotlib.pyplot as plt
import seaborn as sns

# 创建空数据帧
df = pd.DataFrame()

# 添加列
df['x'] = random.sample(range(1, 1000), 5)
df['y'] = random.sample(range(1, 1000), 5)
df['z'] = [1,0,0,1,0]
df['k'] = ['male','male','male','female','female']

# 查看前几行数据
df.head()

	x	y	z	k
0	466	948	1	male
1	832	481	0	male
2	978	465	0	male
3	510	206	1	female
4	848	357	0	female

# 设置散点图样式
sns.set_context("notebook", font_scale=1.1)
sns.set_style("ticks")

# 创建数据帧的散点图
sns.lmplot('x', # 横轴
           'y', # 纵轴
           data=df, # 数据源
           fit_reg=False, # 不要拟合回归直线
           hue="z", # 设置颜色
           scatter_kws={"marker": "D", # 设置标记样式
                        "s": 100}) # 设置标记大小

# 设置标题
plt.title('Histogram of IQ')

# 设置横轴标签
plt.xlabel('Time')

# 设置纵轴标签
plt.ylabel('Deaths')

# <matplotlib.text.Text at 0x112b7bb70>

MatPlotLib 中的分组条形图

%matplotlib inline
import pandas as pd
import matplotlib.pyplot as plt
import numpy as np

raw_data = {'first_name': ['Jason', 'Molly', 'Tina', 'Jake', 'Amy'],
        'pre_score': [4, 24, 31, 2, 3],
        'mid_score': [25, 94, 57, 62, 70],
        'post_score': [5, 43, 23, 23, 51]}
df = pd.DataFrame(raw_data, columns = ['first_name', 'pre_score', 'mid_score', 'post_score'])
df

	first_name	pre_score	mid_score	post_score
0	Jason	4	25	5
1	Molly	24	94	43
2	Tina	31	57	23
3	Jake	2	62	23
4	Amy	3	70	51

# 设置条形的位置和宽度
pos = list(range(len(df['pre_score']))) 
width = 0.25 

# 绘制条形
fig, ax = plt.subplots(figsize=(10,5))

# 使用 pre_score 数据，
# 在位置 pos 上创建条形
plt.bar(pos, 
        # 使用数据 df['pre_score']
        df['pre_score'], 
        # 宽度
        width, 
        # 透明度为 0.5
        alpha=0.5, 
        # 颜色
        color='#EE3224', 
        # 标签是 first_name 的第一个值
        label=df['first_name'][0]) 

# 使用 mid_score 数据，
# 在位置 pos + 一定宽度上创建条形
plt.bar([p + width for p in pos], 
        # 使用数据 df['mid_score']
        df['mid_score'],
        # 宽度
        width, 
        # 透明度为 0.5
        alpha=0.5, 
        # 颜色
        color='#F78F1E', 
        # 标签是 first_name 的第二个值
        label=df['first_name'][1]) 

# 使用 post_score 数据，
# 在位置 pos + 一定宽度上创建条形
plt.bar([p + width*2 for p in pos], 
        # 使用数据 df['post_score']
        df['post_score'], 
        # 宽度
        width, 
        # 透明度为 0.5
        alpha=0.5, 
        # 颜色
        color='#FFC222', 
        # 标签是 first_name 的第三个值
        label=df['first_name'][2]) 

# 设置纵轴标签
ax.set_ylabel('Score')

# 设置标题
ax.set_title('Test Subject Scores')

# 设置 x 刻度的位置
ax.set_xticks([p + 1.5 * width for p in pos])

# 设置 x 刻度的标签
ax.set_xticklabels(df['first_name'])

# 设置横轴和纵轴的区域
plt.xlim(min(pos)-width, max(pos)+width*4)
plt.ylim([0, max(df['pre_score'] + df['mid_score'] + df['post_score'])] )

# 添加图例并展示绘图
plt.legend(['Pre Score', 'Mid Score', 'Post Score'], loc='upper left')
plt.grid()
plt.show()

MatPlotLib 中的直方图

%matplotlib inline
import pandas as pd
import matplotlib.pyplot as plt
import numpy as np
import math

# 设置 ipython 的最大行数
pd.set_option('display.max_row', 1000)

# 将 ipython 的最大列宽设为 50
pd.set_option('display.max_columns', 50)

df = pd.read_csv('https://www.dropbox.com/s/52cb7kcflr8qm2u/5kings_battles_v1.csv?dl=1')
df.head()

	name	year	battle_number	attacker_king	defender_king	attacker_1	attacker_2	attacker_3	attacker_4	defender_1	defender_2	defender_3	defender_4	attacker_outcome	battle_type	major_death	major_capture	attacker_size	defender_size	attacker_commander	defender_commander	summer	location	region	note
0	Battle of the Golden Tooth	298	1	Joffrey/Tommen Baratheon	Robb Stark	Lannister	NaN	NaN	NaN	Tully	NaN	NaN	NaN	win	pitched battle	1	0	15000	4000	Jaime Lannister	Clement Piper, Vance	1	Golden Tooth	The Westerlands	NaN
1	Battle at the Mummer’s Ford	298	2	Joffrey/Tommen Baratheon	Robb Stark	Lannister	NaN	NaN	NaN	Baratheon	NaN	NaN	NaN	win	ambush	1	0	NaN	120	Gregor Clegane	Beric Dondarrion	1	Mummer’s Ford	The Riverlands	NaN
2	Battle of Riverrun	298	3	Joffrey/Tommen Baratheon	Robb Stark	Lannister	NaN	NaN	NaN	Tully	NaN	NaN	NaN	win	pitched battle	0	1	15000	10000	Jaime Lannister, Andros Brax	Edmure Tully, Tytos Blackwood	1	Riverrun	The Riverlands	NaN
—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—
3	Battle of the Green Fork	298	4	Robb Stark	Joffrey/Tommen Baratheon	Stark	NaN	NaN	NaN	Lannister	NaN	NaN	NaN	loss	pitched battle	1	1	18000	20000	Roose Bolton, Wylis Manderly, Medger Cerwyn, H…	Tywin Lannister, Gregor Clegane, Kevan Lannist…	1	Green Fork	The Riverlands	NaN
4	Battle of the Whispering Wood	298	5	Robb Stark	Joffrey/Tommen Baratheon	Stark	Tully	NaN	NaN	Lannister	NaN	NaN	NaN	win	ambush	1	1	1875	6000	Robb Stark, Brynden Tully	Jaime Lannister	1	Whispering Wood	The Riverlands	NaN

# 制作攻击方和防守方大小的两个变量
# 但是当有超过 10000 个攻击方时将其排除在外
data1 = df['attacker_size'][df['attacker_size'] < 90000]
data2 = df['defender_size'][df['attacker_size'] < 90000]

# 创建 2000 个桶
bins = np.arange(data1.min(), data2.max(), 2000) # 固定桶的大小

# 绘制攻击方大小的直方图
plt.hist(data1, 
         bins=bins, 
         alpha=0.5, 
         color='#EDD834',
         label='Attacker')

# 绘制防守方大小的直方图
plt.hist(data2, 
         bins=bins, 
         alpha=0.5, 
         color='#887E43',
         label='Defender')

# 设置图形的 x 和 y 边界
plt.ylim([0, 10])

# 设置标题和标签
plt.title('Histogram of Attacker and Defender Size')
plt.xlabel('Number of troops')
plt.ylabel('Number of battles')
plt.legend(loc='upper right')

plt.show()

# 制作攻击方和防守方大小的两个变量
# 但是当有超过 10000 个攻击方时将其排除在外
data1 = df['attacker_size'][df['attacker_size'] < 90000]
data2 = df['defender_size'][df['attacker_size'] < 90000]

# 创建 10 个桶，最小值为 
# data1 和 data2 的最小值
bins = np.linspace(min(data1 + data2), 
                   # 最大值为它们的最大值
                   max(data1 + data2),
                   # 并分为 10 个桶
                   10)

# 绘制攻击方大小的直方图
plt.hist(data1, 
         # 使用定义好的桶
         bins=bins, 
         # 透明度
         alpha=0.5, 
         # 颜色
         color='#EDD834',
         # 攻击方的标签
         label='Attacker')

# 绘制防守方大小的直方图
plt.hist(data2, 
         # 使用定义好的桶
         bins=bins, 
         # 透明度
         alpha=0.5, 
         # 颜色
         color='#887E43',
         # 防守方的标签
         label='Defender')

# 设置图形的 x 和 y 边界
plt.ylim([0, 10])

# 设置标题和标签
plt.title('Histogram of Attacker and Defender Size')
plt.xlabel('Number of troops')
plt.ylabel('Number of battles')
plt.legend(loc='upper right')

plt.show()

从 Pandas 数据帧生成 MatPlotLib 散点图

%matplotlib inline
import pandas as pd
import matplotlib.pyplot as plt
import numpy as np

raw_data = {'first_name': ['Jason', 'Molly', 'Tina', 'Jake', 'Amy'], 
        'last_name': ['Miller', 'Jacobson', 'Ali', 'Milner', 'Cooze'], 
        'female': [0, 1, 1, 0, 1],
        'age': [42, 52, 36, 24, 73], 
        'preTestScore': [4, 24, 31, 2, 3],
        'postTestScore': [25, 94, 57, 62, 70]}
df = pd.DataFrame(raw_data, columns = ['first_name', 'last_name', 'age', 'female', 'preTestScore', 'postTestScore'])
df

	first_name	last_name	age	female	preTestScore	postTestScore
0	Jason	Miller	42	0	4	25
1	Molly	Jacobson	52	1	24	94
2	Tina	Ali	36	1	31	57
3	Jake	Milner	24	0	2	62
4	Amy	Cooze	73	1	3	70

# preTestScore 和 postTestScore 的散点图
# 每个点的大小取决于年龄
plt.scatter(df.preTestScore, df.postTestScore
, s=df.age)

# <matplotlib.collections.PathCollection at 0x10ca42b00>

# preTestScore 和 postTestScore 的散点图
# 大小为 300，颜色取决于性别
plt.scatter(df.preTestScore, df.postTestScore, s=300, c=df.female)

# <matplotlib.collections.PathCollection at 0x10cb90a90>

Matplotlib 的简单示例

# 让 Jupyter 加载 matplotlib 
# 并内联创建所有绘图（也就是在页面上）
%matplotlib inline

import matplotlib.pyplot as pyplot

pyplot.plot([1.6, 2.7])

# [<matplotlib.lines.Line2D at 0x10c4e7978>]

MatPlotLib 中的饼图

%matplotlib inline
import pandas as pd
import matplotlib.pyplot as plt

raw_data = {'officer_name': ['Jason', 'Molly', 'Tina', 'Jake', 'Amy'],
        'jan_arrests': [4, 24, 31, 2, 3],
        'feb_arrests': [25, 94, 57, 62, 70],
        'march_arrests': [5, 43, 23, 23, 51]}
df = pd.DataFrame(raw_data, columns = ['officer_name', 'jan_arrests', 'feb_arrests', 'march_arrests'])
df

	officer_name	jan_arrests	feb_arrests	march_arrests
0	Jason	4	25	5
1	Molly	24	94	43
2	Tina	31	57	23
3	Jake	2	62	23
4	Amy	3	70	51

# 创建一列，其中包含每个官员的总逮捕数
df['total_arrests'] = df['jan_arrests'] + df['feb_arrests'] + df['march_arrests']
df

	officer_name	jan_arrests	feb_arrests	march_arrests	total_arrests
0	Jason	4	25	5	34
1	Molly	24	94	43	161
2	Tina	31	57	23	111
3	Jake	2	62	23	87
4	Amy	3	70	51	124

# （从 iWantHue）创建一列颜色
colors = ["#E13F29", "#D69A80", "#D63B59", "#AE5552", "#CB5C3B", "#EB8076", "#96624E"]

# 创建饼图
plt.pie(
    # 使用数据 total_arrests
    df['total_arrests'],
    # 标签为官员名称
    labels=df['officer_name'],
    # 没有阴影
    shadow=False,
    # 颜色
    colors=colors,
    # 将一块扇形移出去
    explode=(0, 0, 0, 0, 0.15),
    # 起始角度为 90 度
    startangle=90,
    # 将百分比列为分数
    autopct='%1.1f%%',
    )

# 使饼状图为正圆
plt.axis('equal')

# 查看绘图
plt.tight_layout()
plt.show()

MatPlotLib 中的散点图

%matplotlib inline
import pandas as pd
import matplotlib.pyplot as plt
import numpy as np

# 展示 ipython 的最大行数
pd.set_option('display.max_row', 1000)

# 将 ipython 的最大列宽设为 50
pd.set_option('display.max_columns', 50)

df = pd.read_csv('https://raw.githubusercontent.com/chrisalbon/war_of_the_five_kings_dataset/master/5kings_battles_v1.csv')
df.head()

	name	year	battle_number	attacker_king	defender_king	attacker_1	attacker_2	attacker_3	attacker_4	defender_1	defender_2	defender_3	defender_4	attacker_outcome	battle_type	major_death	major_capture	attacker_size	defender_size	attacker_commander	defender_commander	summer	location	region	note
0	Battle of the Golden Tooth	298	1	Joffrey/Tommen Baratheon	Robb Stark	Lannister	NaN	NaN	NaN	Tully	NaN	NaN	NaN	win	pitched battle	1.0	0.0	15000.0	4000.0	Jaime Lannister	Clement Piper, Vance	1.0	Golden Tooth	The Westerlands	NaN
1	Battle at the Mummer’s Ford	298	2	Joffrey/Tommen Baratheon	Robb Stark	Lannister	NaN	NaN	NaN	Baratheon	NaN	NaN	NaN	win	ambush	1.0	0.0	NaN	120.0	Gregor Clegane	Beric Dondarrion	1.0	Mummer’s Ford	The Riverlands	NaN
2	Battle of Riverrun	298	3	Joffrey/Tommen Baratheon	Robb Stark	Lannister	NaN	NaN	NaN	Tully	NaN	NaN	NaN	win	pitched battle	0.0	1.0	15000.0	10000.0	Jaime Lannister, Andros Brax	Edmure Tully, Tytos Blackwood	1.0	Riverrun	The Riverlands	NaN
3	Battle of the Green Fork	298	4	Robb Stark	Joffrey/Tommen Baratheon	Stark	NaN	NaN	NaN	Lannister	NaN	NaN	NaN	loss	pitched battle	1.0	1.0	18000.0	20000.0	Roose Bolton, Wylis Manderly, Medger Cerwyn, H…	Tywin Lannister, Gregor Clegane, Kevan Lannist…	1.0	Green Fork	The Riverlands	NaN
4	Battle of the Whispering Wood	298	5	Robb Stark	Joffrey/Tommen Baratheon	Stark	Tully	NaN	NaN	Lannister	NaN	NaN	NaN	win	ambush	1.0	1.0	1875.0	6000.0	Robb Stark, Brynden Tully	Jaime Lannister	1.0	Whispering Wood	The Riverlands	NaN

# 创建图形
plt.figure(figsize=(10,8))

# 创建散点图
            # 298 年的攻击方大小为 x 轴
plt.scatter(df['attacker_size'][df['year'] == 298], 
            # 298 年的防守方大小为 y 轴
            df['defender_size'][df['year'] == 298], 
            # 标记
            marker='x', 
            # 颜色
            color='b',
            # 透明度
            alpha=0.7,
            # 大小
            s = 124,
            # 标签
            label='Year 298')

            # 299 年的攻击方大小为 x 轴
plt.scatter(df['attacker_size'][df['year'] == 299], 
            # 299 年的防守方大小为 y 轴
            df['defender_size'][df['year'] == 299], 
            # 标记
            marker='o', 
            # 颜色
            color='r', 
            # 透明度
            alpha=0.7,
            # 大小
            s = 124,
            # 标签
            label='Year 299')

            # 300 年的攻击方大小为 x 轴
plt.scatter(df['attacker_size'][df['year'] == 300], 
            # 300 年的防守方大小为 x 轴
            df['defender_size'][df['year'] == 300], 
            # 标记
            marker='^', 
            # 颜色
            color='g', 
            # 透明度
            alpha=0.7, 
            # 大小
            s = 124,
            # 标签
            label='Year 300')

# 标题
plt.title('Battles Of The War Of The Five Kings')

# y 标签
plt.ylabel('Defender Size')

# x 标签
plt.xlabel('Attacker Size')

# 图例
plt.legend(loc='upper right')

# 设置图形边界
plt.xlim([min(df['attacker_size'])-1000, max(df['attacker_size'])+1000])
plt.ylim([min(df['defender_size'])-1000, max(df['defender_size'])+1000])

plt.show()

MatPlotLib 中的栈式百分比条形图

%matplotlib inline
import pandas as pd
import matplotlib.pyplot as plt

raw_data = {'first_name': ['Jason', 'Molly', 'Tina', 'Jake', 'Amy'],
        'pre_score': [4, 24, 31, 2, 3],
        'mid_score': [25, 94, 57, 62, 70],
        'post_score': [5, 43, 23, 23, 51]}
df = pd.DataFrame(raw_data, columns = ['first_name', 'pre_score', 'mid_score', 'post_score'])
df

	first_name	pre_score	mid_score	post_score
0	Jason	4	25	5
1	Molly	24	94	43
2	Tina	31	57	23
3	Jake	2	62	23
4	Amy	3	70	51

# 创建带有一个子图的图形
f, ax = plt.subplots(1, figsize=(10,5))

# 将条宽设为 1
bar_width = 1

# 条形左边界的位置
bar_l = [i for i in range(len(df['pre_score']))] 

# x 轴刻度的位置（条形的中心是条形标签）
tick_pos = [i+(bar_width/2) for i in bar_l] 

# 创建每个参与者的总得分
totals = [i+j+k for i,j,k in zip(df['pre_score'], df['mid_score'], df['post_score'])]

# 创建每个参与者的 pre_score 和总得分的百分比
pre_rel = [i / j * 100 for  i,j in zip(df['pre_score'], totals)]

# 创建每个参与者的 mid_score 和总得分的百分比
mid_rel = [i / j * 100 for  i,j in zip(df['mid_score'], totals)]

# 创建每个参与者的 post_score 和总得分的百分比
post_rel = [i / j * 100 for  i,j in zip(df['post_score'], totals)]

# 在位置 bar_1 创建条形图
ax.bar(bar_l, 
       # 使用数据 pre_rel
       pre_rel, 
       # 标签 
       label='Pre Score', 
       # 透明度
       alpha=0.9, 
       # 颜色
       color='#019600',
       # 条形宽度
       width=bar_width,
       # 边框颜色
       edgecolor='white'
       )

# 在位置 bar_1 创建条形图
ax.bar(bar_l, 
       # 使用数据 mid_rel
       mid_rel, 
       # 底部为 pre_rel
       bottom=pre_rel, 
       # 标签
       label='Mid Score', 
       # 透明度
       alpha=0.9, 
       # 颜色
       color='#3C5F5A', 
       # 条形宽度
       width=bar_width,
       # 边框颜色
       edgecolor='white'
       )

# Create a bar chart in position bar_1
ax.bar(bar_l, 
       # 使用数据 post_rel
       post_rel, 
       # 底部为 pre_rel 和 mid_rel
       bottom=[i+j for i,j in zip(pre_rel, mid_rel)], 
       # 标签
       label='Post Score',
       # 透明度
       alpha=0.9, 
       # 颜色
       color='#219AD8', 
       # 条形宽度
       width=bar_width,
       # 边框颜色
       edgecolor='white'
       )

# 将刻度设为 first_name
plt.xticks(tick_pos, df['first_name'])
ax.set_ylabel("Percentage")
ax.set_xlabel("")

# 设置图形边界
plt.xlim([min(tick_pos)-bar_width, max(tick_pos)+bar_width])
plt.ylim(-10, 110)

# 旋转轴标签
plt.setp(plt.gca().get_xticklabels(), rotation=45, horizontalalignment='right')

# 展示绘图
plt.show()