数据分析——从入门到精通(十三)

import numpy as np
import pandas as pd 
from pandas import Series,DataFrame

import matplotlib.pyplot as plt

# 当前文档显示matplotlib的图像：可视化
%matplotlib inline

s1 = Series(np.random.randint(100,150,size=20))
s1

0     127
1     124
2     143
3     139
4     139
5     145
6     147
7     132
8     106
9     117
10    123
11    136
12    113
13    118
14    141
15    131
16    109
17    149
18    136
19    101
dtype: int32

s1.plot()

<matplotlib.axes._subplots.AxesSubplot at 0x22ed370>

df1 = DataFrame(np.random.randint(0,100,size=(20,3)),columns=['SQLite','Oracle','MySQL'])
df1

	SQLite	Oracle	MySQL
0	45	54	9
1	52	63	0
2	44	97	35
3	49	51	95
4	35	51	27
5	50	80	73
6	89	17	22
7	87	32	15
8	32	93	60
9	54	12	54
10	7	93	21
11	81	99	81
12	32	80	47
13	42	34	27
14	57	3	98
15	49	0	78
16	15	66	13
17	65	66	5
18	45	7	99
19	53	98	24

df1.plot()

<matplotlib.axes._subplots.AxesSubplot at 0x1878550>

# 按照SQLite排序
df1.sort_values('SQLite')

	SQLite	Oracle	MySQL
10	7	93	21
16	15	66	13
8	32	93	60
12	32	80	47
4	35	51	27
13	42	34	27
2	44	97	35
0	45	54	9
18	45	7	99
3	49	51	95
15	49	0	78
5	50	80	73
1	52	63	0
19	53	98	24
9	54	12	54
14	57	3	98
17	65	66	5
11	81	99	81
7	87	32	15
6	89	17	22

df1.sort_values('SQLite',ascending=False)

	SQLite	Oracle	MySQL
6	89	17	22
7	87	32	15
11	81	99	81
17	65	66	5
14	57	3	98
9	54	12	54
19	53	98	24
1	52	63	0
5	50	80	73
3	49	51	95
15	49	0	78
18	45	7	99
0	45	54	9
2	44	97	35
13	42	34	27
4	35	51	27
12	32	80	47
8	32	93	60
16	15	66	13
10	7	93	21

df1.sort_values('SQLite',ascending=False).plot()

<matplotlib.axes._subplots.AxesSubplot at 0x1907cd0>

df1.sort_values('SQLite',ascending=False).reset_index(drop=True).plot()

<matplotlib.axes._subplots.AxesSubplot at 0x6862c30>

# 多列排序
df1.sort_values(['SQLite','Oracle'],ascending=False).reset_index(drop=True).plot()

<matplotlib.axes._subplots.AxesSubplot at 0x7ba0eb0>

df1.sort_values(['SQLite','Oracle'],ascending=False).reset_index(drop=True)[['SQLite','Oracle']].plot()

<matplotlib.axes._subplots.AxesSubplot at 0x9598870>

df1.plot()     # 绘制线型图，返回的是一个子画布对象

<matplotlib.axes._subplots.AxesSubplot at 0x962ae10>

柱状图

• kind=‘bar’ 垂直方向
• kind=‘barh’ 水平方向

s1.plot()

<matplotlib.axes._subplots.AxesSubplot at 0x969f7f0>

s1.plot(kind='bar')

<matplotlib.axes._subplots.AxesSubplot at 0x975faf0>

s1.plot(kind='bar',colormap='rainbow')

<matplotlib.axes._subplots.AxesSubplot at 0x9842dd0>

from matplotlib.colors import ListedColormap
# 自定义加三种颜色
cm1 = ListedColormap([np.random.uniform(0,1,size=3) for i in range(3)])
# 自定义加三种颜色
cm2 = ListedColormap(['red','blue','green'])

s1.plot(kind='bar',colormap=cm1)

<matplotlib.axes._subplots.AxesSubplot at 0xa1a1bb0>

s1.plot(kind='bar',colormap=cm2)

<matplotlib.axes._subplots.AxesSubplot at 0xa27f190>

df1

	SQLite	Oracle	MySQL
0	45	54	9
1	52	63	0
2	44	97	35
3	49	51	95
4	35	51	27
5	50	80	73
6	89	17	22
7	87	32	15
8	32	93	60
9	54	12	54
10	7	93	21
11	81	99	81
12	32	80	47
13	42	34	27
14	57	3	98
15	49	0	78
16	15	66	13
17	65	66	5
18	45	7	99
19	53	98	24

df1.plot(kind='bar')

<matplotlib.axes._subplots.AxesSubplot at 0xa3a1bb0>

df1.plot(kind='bar',colormap=cm1)

<matplotlib.axes._subplots.AxesSubplot at 0xb4ece90>

df1.plot(kind='bar',colormap=cm2)

<matplotlib.axes._subplots.AxesSubplot at 0xb5a4650>

# 解决matplotlib的中文乱码问题
plt.rcParams['font.sans-serif'] = ['SimHei']
# 解决坐标轴的刻度乱码问题
plt.rcParams['axes.unicode_minus'] = False

df1.plot(kind='bar',colormap=cm2,title='20位同学的期中考试成绩')

<matplotlib.axes._subplots.AxesSubplot at 0xb7c54b0>

df1.plot(kind='bar',colormap=cm2,title='20位同学的期中考试成绩',fontsize=20)

<matplotlib.axes._subplots.AxesSubplot at 0xb9360b0>

ax = df1.plot(kind='bar',colormap=cm2,title='20位同学的期中考试成绩',fontsize=20)

# 判断ax是否有"set_title"的属性
hasattr(ax,'set_title')

True

# 判断ax里面是否有"title"的属性
hasattr(ax,'title')

True

ax.title

Text(0.5, 1, '20位同学的期中考试成绩')

# 查看ax.title的position的位置
ax.title.get_position()

(0.5, 1)

# 查看ax.title的所有属性
dir(ax.title)

['__class__',
 '__delattr__',
 '__dict__',
 '__dir__',
 '__doc__',
 '__eq__',
 '__format__',
 '__ge__',
 '__getattribute__',
 '__getstate__',
 '__gt__',
 '__hash__',
 '__init__',
 '__init_subclass__',
 '__le__',
 '__lt__',
 '__module__',
 '__ne__',
 '__new__',
 '__reduce__',
 '__reduce_ex__',
 '__repr__',
 '__setattr__',
 '__sizeof__',
 '__str__',
 '__subclasshook__',
 '__weakref__',
 '_agg_filter',
 '_alias_map',
 '_alpha',
 '_animated',
 '_axes',
 '_bbox_patch',
 '_cached',
 '_clipon',
 '_clippath',
 '_color',
 '_contains',
 '_draw_bbox',
 '_fontproperties',
 '_get_clipping_extent_bbox',
 '_get_dist_to_box',
 '_get_layout',
 '_get_multialignment',
 '_get_rendered_text_width',
 '_get_wrap_line_width',
 '_get_wrapped_text',
 '_get_xy_display',
 '_gid',
 '_horizontalalignment',
 '_in_layout',
 '_label',
 '_linespacing',
 '_mouseover',
 '_multialignment',
 '_oid',
 '_path_effects',
 '_picker',
 '_preprocess_math',
 '_prop_order',
 '_propobservers',
 '_rasterized',
 '_remove_method',
 '_renderer',
 '_rotation',
 '_rotation_mode',
 '_set_gc_clip',
 '_sketch',
 '_snap',
 '_stale',
 '_sticky_edges',
 '_text',
 '_transform',
 '_transformSet',
 '_update_clip_properties',
 '_url',
 '_usetex',
 '_verticalalignment',
 '_visible',
 '_wrap',
 '_x',
 '_y',
 'add_callback',
 'aname',
 'axes',
 'clipbox',
 'contains',
 'convert_xunits',
 'convert_yunits',
 'draw',
 'eventson',
 'figure',
 'findobj',
 'format_cursor_data',
 'get_agg_filter',
 'get_alpha',
 'get_animated',
 'get_bbox_patch',
 'get_c',
 'get_children',
 'get_clip_box',
 'get_clip_on',
 'get_clip_path',
 'get_color',
 'get_contains',
 'get_cursor_data',
 'get_family',
 'get_figure',
 'get_font_properties',
 'get_fontfamily',
 'get_fontname',
 'get_fontproperties',
 'get_fontsize',
 'get_fontstyle',
 'get_fontvariant',
 'get_fontweight',
 'get_gid',
 'get_ha',
 'get_horizontalalignment',
 'get_in_layout',
 'get_label',
 'get_name',
 'get_path_effects',
 'get_picker',
 'get_position',
 'get_prop_tup',
 'get_rasterized',
 'get_rotation',
 'get_rotation_mode',
 'get_size',
 'get_sketch_params',
 'get_snap',
 'get_stretch',
 'get_style',
 'get_text',
 'get_tightbbox',
 'get_transform',
 'get_transformed_clip_path_and_affine',
 'get_unitless_position',
 'get_url',
 'get_usetex',
 'get_va',
 'get_variant',
 'get_verticalalignment',
 'get_visible',
 'get_weight',
 'get_window_extent',
 'get_wrap',
 'get_zorder',
 'have_units',
 'is_math_text',
 'is_transform_set',
 'mouseover',
 'pchanged',
 'pick',
 'pickable',
 'properties',
 'remove',
 'remove_callback',
 'set',
 'set_agg_filter',
 'set_alpha',
 'set_animated',
 'set_backgroundcolor',
 'set_bbox',
 'set_c',
 'set_clip_box',
 'set_clip_on',
 'set_clip_path',
 'set_color',
 'set_contains',
 'set_family',
 'set_figure',
 'set_font_properties',
 'set_fontfamily',
 'set_fontname',
 'set_fontproperties',
 'set_fontsize',
 'set_fontstretch',
 'set_fontstyle',
 'set_fontvariant',
 'set_fontweight',
 'set_gid',
 'set_ha',
 'set_horizontalalignment',
 'set_in_layout',
 'set_label',
 'set_linespacing',
 'set_ma',
 'set_multialignment',
 'set_name',
 'set_path_effects',
 'set_picker',
 'set_position',
 'set_rasterized',
 'set_rotation',
 'set_rotation_mode',
 'set_size',
 'set_sketch_params',
 'set_snap',
 'set_stretch',
 'set_style',
 'set_text',
 'set_transform',
 'set_url',
 'set_usetex',
 'set_va',
 'set_variant',
 'set_verticalalignment',
 'set_visible',
 'set_weight',
 'set_wrap',
 'set_x',
 'set_y',
 'set_zorder',
 'stale',
 'stale_callback',
 'sticky_edges',
 'update',
 'update_bbox_position_size',
 'update_from',
 'zorder']

# 去掉以"_"开头的属性
[attr for attr in dir(ax.title) if not attr.startswith('_')]

['add_callback',
 'aname',
 'axes',
 'clipbox',
 'contains',
 'convert_xunits',
 'convert_yunits',
 'draw',
 'eventson',
 'figure',
 'findobj',
 'format_cursor_data',
 'get_agg_filter',
 'get_alpha',
 'get_animated',
 'get_bbox_patch',
 'get_c',
 'get_children',
 'get_clip_box',
 'get_clip_on',
 'get_clip_path',
 'get_color',
 'get_contains',
 'get_cursor_data',
 'get_family',
 'get_figure',
 'get_font_properties',
 'get_fontfamily',
 'get_fontname',
 'get_fontproperties',
 'get_fontsize',
 'get_fontstyle',
 'get_fontvariant',
 'get_fontweight',
 'get_gid',
 'get_ha',
 'get_horizontalalignment',
 'get_in_layout',
 'get_label',
 'get_name',
 'get_path_effects',
 'get_picker',
 'get_position',
 'get_prop_tup',
 'get_rasterized',
 'get_rotation',
 'get_rotation_mode',
 'get_size',
 'get_sketch_params',
 'get_snap',
 'get_stretch',
 'get_style',
 'get_text',
 'get_tightbbox',
 'get_transform',
 'get_transformed_clip_path_and_affine',
 'get_unitless_position',
 'get_url',
 'get_usetex',
 'get_va',
 'get_variant',
 'get_verticalalignment',
 'get_visible',
 'get_weight',
 'get_window_extent',
 'get_wrap',
 'get_zorder',
 'have_units',
 'is_math_text',
 'is_transform_set',
 'mouseover',
 'pchanged',
 'pick',
 'pickable',
 'properties',
 'remove',
 'remove_callback',
 'set',
 'set_agg_filter',
 'set_alpha',
 'set_animated',
 'set_backgroundcolor',
 'set_bbox',
 'set_c',
 'set_clip_box',
 'set_clip_on',
 'set_clip_path',
 'set_color',
 'set_contains',
 'set_family',
 'set_figure',
 'set_font_properties',
 'set_fontfamily',
 'set_fontname',
 'set_fontproperties',
 'set_fontsize',
 'set_fontstretch',
 'set_fontstyle',
 'set_fontvariant',
 'set_fontweight',
 'set_gid',
 'set_ha',
 'set_horizontalalignment',
 'set_in_layout',
 'set_label',
 'set_linespacing',
 'set_ma',
 'set_multialignment',
 'set_name',
 'set_path_effects',
 'set_picker',
 'set_position',
 'set_rasterized',
 'set_rotation',
 'set_rotation_mode',
 'set_size',
 'set_sketch_params',
 'set_snap',
 'set_stretch',
 'set_style',
 'set_text',
 'set_transform',
 'set_url',
 'set_usetex',
 'set_va',
 'set_variant',
 'set_verticalalignment',
 'set_visible',
 'set_weight',
 'set_wrap',
 'set_x',
 'set_y',
 'set_zorder',
 'stale',
 'stale_callback',
 'sticky_edges',
 'update',
 'update_bbox_position_size',
 'update_from',
 'zorder']

# 调整标题的字体大小
ax = df1.plot(kind='bar',colormap=cm2,title='20位同学的期中考试成绩',fontsize=20)
ax.title.set_size(20)
plt.show()

ax = df1.plot(kind='bar',colormap=cm2,title='20位同学的期中考试成绩',fontsize=20)
ax.title.set_size(20)
ax.title.set_position((0.5,1,1))  # 0.5指居中的位置
plt.show()

fig,axes = plt.subplots(2,1,figsize=(8,10))   # figsize=(8,10) 画布大小
s1.plot(kind='bar',ax=axes[0],title='图1')   # 插入子画布1
s1.plot(kind='barh',ax=axes[1],title='图2')  # 插入子画布2
plt.show()    # 显示绘制的画布

df1.plot(kind='bar')

<matplotlib.axes._subplots.AxesSubplot at 0xbfe0930>

# 设置画布大小
fig, axes = plt.subplots(figsize=(10,6))
df1.plot(kind='bar',ax=axes)   # 只有一个子画布时，不需要索引操作

<matplotlib.axes._subplots.AxesSubplot at 0xc22da50>

# 设置画布大小
fig, axes = plt.subplots(figsize=(10,6))
# stacked 堆栈的方式，bool值
df1.plot(kind='bar',ax=axes,stacked=True)   # 只有一个子画布时，不需要索引操作

<matplotlib.axes._subplots.AxesSubplot at 0xbb404b0>

直方图或密度图

• .hist(bins=10,alpha=0.5,color=“gray”,normed|density=True)
• .plot(kind=kde)

s1.hist()

<matplotlib.axes._subplots.AxesSubplot at 0xbbedc30>

s1.hist(bins=20) # bins=20 柱子放20个

<matplotlib.axes._subplots.AxesSubplot at 0xbbb66b0>

s1.hist(bins=2)

<matplotlib.axes._subplots.AxesSubplot at 0xc495690>

s1.sort_values()

19    101
8     106
16    109
12    113
9     117
13    118
10    123
1     124
0     127
15    131
7     132
18    136
11    136
3     139
4     139
14    141
2     143
5     145
6     147
17    149
dtype: int32

s2 = Series(np.random.randn(100))
s2.hist()

<matplotlib.axes._subplots.AxesSubplot at 0xc535c30>

s2 = Series(np.random.randn(100))
s2.hist(bins=50)

<matplotlib.axes._subplots.AxesSubplot at 0xc58ca90>

s2 = Series(np.random.randn(100))
s2.hist(bins=50)
s2.plot(kind='kde',color='red')

<matplotlib.axes._subplots.AxesSubplot at 0xd720070>

s2 = Series(np.random.randn(100))
s2.hist(bins=50,alpha=0.3,density=True)   # 将直方的数值转化为密度值
s2.plot(kind='kde',color='red')

<matplotlib.axes._subplots.AxesSubplot at 0xc3ad670>

散点图

• plt.scatter(x,y,marksize=50,color=)
• dataframe.plot(x列，y列，kind=‘scatter’)
• pd.plotting.scatter_maxtrix(frame,ax,diagonal=‘hist|kde’,s=50,c=‘’) 散点矩阵图

df1

	SQLite	Oracle	MySQL
0	45	54	9
1	52	63	0
2	44	97	35
3	49	51	95
4	35	51	27
5	50	80	73
6	89	17	22
7	87	32	15
8	32	93	60
9	54	12	54
10	7	93	21
11	81	99	81
12	32	80	47
13	42	34	27
14	57	3	98
15	49	0	78
16	15	66	13
17	65	66	5
18	45	7	99
19	53	98	24

# 查看SQLite,Oracle的关系
df1.plot('SQLite','Oracle')

<matplotlib.axes._subplots.AxesSubplot at 0xd7b91b0>

# 查看SQLite,Oracle的关系,散点图
df1.plot('SQLite','Oracle',kind='scatter')

<matplotlib.axes._subplots.AxesSubplot at 0xd7fae90>

# 查看Oracle,MySQL的关系,散点图
df1.plot('Oracle','MySQL',kind='scatter')

<matplotlib.axes._subplots.AxesSubplot at 0xd832c50>

plt.scatter(df1.MySQL,df1.Oracle)

<matplotlib.collections.PathCollection at 0xd8af3f0>

plt.scatter(df1.MySQL,df1.Oracle,s=100)

<matplotlib.collections.PathCollection at 0xd7fa090>

plt.scatter(df1.MySQL,df1.Oracle,s=100,c='r')  # s指大小，c指颜色

<matplotlib.collections.PathCollection at 0xd995bb0>

plt.scatter(df1.MySQL,df1.Oracle,s=100,c='k')  # c='k' 中 k ----->blank 黑色

<matplotlib.collections.PathCollection at 0xda0f6d0>

pd.plotting.scatter_matrix(df1,diagonal='hist') # diagonal='hist'直方图

array([[<matplotlib.axes._subplots.AxesSubplot object at 0x0DA2A610>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x0DC02AB0>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x0DC26B50>],
       [<matplotlib.axes._subplots.AxesSubplot object at 0x0DC45C30>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x0DC66D10>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x0DC86DF0>],
       [<matplotlib.axes._subplots.AxesSubplot object at 0x0DCA6ED0>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x0DCC8FB0>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x0DCD3570>]],
      dtype=object)

pd.plotting.scatter_matrix(df1,diagonal='kde')  # diagonal='kde'密度图

array([[<matplotlib.axes._subplots.AxesSubplot object at 0x0DA1A930>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x0E58BA10>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x0E5A8AF0>],
       [<matplotlib.axes._subplots.AxesSubplot object at 0x0E5C8BD0>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x0E5E8CB0>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x0E609D90>],
       [<matplotlib.axes._subplots.AxesSubplot object at 0x0E62AEF0>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x0E64C9B0>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x0E64CF50>]],
      dtype=object)

pd.plotting.scatter_matrix(df1,diagonal='kde',c='r')

array([[<matplotlib.axes._subplots.AxesSubplot object at 0x0E578E10>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x0F6F2BB0>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x0F70EC90>],
       [<matplotlib.axes._subplots.AxesSubplot object at 0x0F72FD70>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x0F74EE50>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x0F76FF30>],
       [<matplotlib.axes._subplots.AxesSubplot object at 0x0F78FA70>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x0F7B1B50>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x0F7BC0F0>]],
      dtype=object)

pd.plotting.scatter_matrix(df1,diagonal='kde',c='r',s=80)

array([[<matplotlib.axes._subplots.AxesSubplot object at 0x0E6E0CD0>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x0F87AD90>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x0F8A4E50>],
       [<matplotlib.axes._subplots.AxesSubplot object at 0x0F8C2F30>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x0F8E3A70>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x0F902B50>],
       [<matplotlib.axes._subplots.AxesSubplot object at 0x0F925C30>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x0F944D10>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x0F94F2D0>]],
      dtype=object)