pandas简单教程

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

对象创建

通过传入一些值的列表来创建一个Series， Pandas会自动创建一个默认的整数索引：

s = pd.Series([1,3,5,np.nan,6,8])
s

0    1.0
1    3.0
2    5.0
3    NaN
4    6.0
5    8.0
dtype: float64

通过传递带有日期时间索引和带标签列的NumPy数组来创建DataFrame：

dates = pd.date_range('20130101',periods=6)
dates
df = pd.DataFrame(np.random.randn(6,4),index=dates,columns=list('ABCD'))
df

	A	B	C	D
2013-01-01	-0.828948	0.281765	0.803692	0.030016
2013-01-02	0.418212	1.537528	0.407742	0.625449
2013-01-03	0.746757	-0.338140	-0.734583	-2.377116
2013-01-04	-0.507705	-0.409561	-2.596286	0.464993
2013-01-05	-0.154101	-0.675057	-0.747016	-0.192082
2013-01-06	0.892789	-1.848313	0.897434	0.157656

通过传递可以转化为类似Series的dict对象来创建DataFrame:

df2 = pd.DataFrame({ 'A' : 1.,
                    'B' : pd.Timestamp('20130102'),
                    'C' : pd.Series(1,index=list(range(4)),dtype='float32'),
                    'D' : np.array([3] * 4,dtype='int32'),
                     'E' : pd.Categorical(["test","train","test","train"]),
                     'F' : 'foo' })
df2

	A	B	C	D	E	F
0	1.0	2013-01-02	1.0	3	test	foo
1	1.0	2013-01-02	1.0	3	train	foo
2	1.0	2013-01-02	1.0	3	test	foo
3	1.0	2013-01-02	1.0	3	train	foo

df2.dtypes

A           float64
B    datetime64[ns]
C           float32
D             int32
E          category
F            object
dtype: object

查看数据

df.head()

	A	B	C	D
2013-01-01	-0.828948	0.281765	0.803692	0.030016
2013-01-02	0.418212	1.537528	0.407742	0.625449
2013-01-03	0.746757	-0.338140	-0.734583	-2.377116
2013-01-04	-0.507705	-0.409561	-2.596286	0.464993
2013-01-05	-0.154101	-0.675057	-0.747016	-0.192082

df.tail(2)

	A	B	C	D
2013-01-05	-0.154101	-0.675057	-0.747016	-0.192082
2013-01-06	0.892789	-1.848313	0.897434	0.157656

显示索引、列和底层NumPy数据：

df.index

DatetimeIndex(['2013-01-01', '2013-01-02', '2013-01-03', '2013-01-04',
               '2013-01-05', '2013-01-06'],
              dtype='datetime64[ns]', freq='D')

df.columns

Index(['A', 'B', 'C', 'D'], dtype='object')

df.values

array([[-0.82894761,  0.28176527,  0.80369199,  0.03001636],
       [ 0.41821203,  1.53752828,  0.40774162,  0.62544912],
       [ 0.74675688, -0.33814015, -0.73458287, -2.3771161 ],
       [-0.5077046 , -0.4095612 , -2.59628619,  0.46499331],
       [-0.15410053, -0.67505665, -0.74701636, -0.19208195],
       [ 0.89278944, -1.84831322,  0.8974336 ,  0.15765575]])

df.describe()#显示数据的快速统计摘要

	A	B	C	D
count	6.000000	6.000000	6.000000	6.000000
mean	0.094501	-0.241963	-0.328170	-0.215181
std	0.699243	1.117690	1.327384	1.099356
min	-0.828948	-1.848313	-2.596286	-2.377116
25%	-0.419304	-0.608683	-0.743908	-0.136557
50%	0.132056	-0.373851	-0.163421	0.093836
75%	0.664621	0.126789	0.704704	0.388159
max	0.892789	1.537528	0.897434	0.625449

df.T

	2013-01-01 00:00:00	2013-01-02 00:00:00	2013-01-03 00:00:00	2013-01-04 00:00:00	2013-01-05 00:00:00	2013-01-06 00:00:00
A	-0.828948	0.418212	0.746757	-0.507705	-0.154101	0.892789
B	0.281765	1.537528	-0.338140	-0.409561	-0.675057	-1.848313
C	0.803692	0.407742	-0.734583	-2.596286	-0.747016	0.897434
D	0.030016	0.625449	-2.377116	0.464993	-0.192082	0.157656

print( df.sort_index(axis=1, ascending=False))
print( df.sort_index(axis=0, ascending=False))
print( df.sort_index(axis=1, ascending=True))
print( df.sort_index(axis=0, ascending=True))

                   D         C         B         A
2013-01-01  0.030016  0.803692  0.281765 -0.828948
2013-01-02  0.625449  0.407742  1.537528  0.418212
2013-01-03 -2.377116 -0.734583 -0.338140  0.746757
2013-01-04  0.464993 -2.596286 -0.409561 -0.507705
2013-01-05 -0.192082 -0.747016 -0.675057 -0.154101
2013-01-06  0.157656  0.897434 -1.848313  0.892789
                   A         B         C         D
2013-01-06  0.892789 -1.848313  0.897434  0.157656
2013-01-05 -0.154101 -0.675057 -0.747016 -0.192082
2013-01-04 -0.507705 -0.409561 -2.596286  0.464993
2013-01-03  0.746757 -0.338140 -0.734583 -2.377116
2013-01-02  0.418212  1.537528  0.407742  0.625449
2013-01-01 -0.828948  0.281765  0.803692  0.030016
                   A         B         C         D
2013-01-01 -0.828948  0.281765  0.803692  0.030016
2013-01-02  0.418212  1.537528  0.407742  0.625449
2013-01-03  0.746757 -0.338140 -0.734583 -2.377116
2013-01-04 -0.507705 -0.409561 -2.596286  0.464993
2013-01-05 -0.154101 -0.675057 -0.747016 -0.192082
2013-01-06  0.892789 -1.848313  0.897434  0.157656
                   A         B         C         D
2013-01-01 -0.828948  0.281765  0.803692  0.030016
2013-01-02  0.418212  1.537528  0.407742  0.625449
2013-01-03  0.746757 -0.338140 -0.734583 -2.377116
2013-01-04 -0.507705 -0.409561 -2.596286  0.464993
2013-01-05 -0.154101 -0.675057 -0.747016 -0.192082
2013-01-06  0.892789 -1.848313  0.897434  0.157656

 df.sort_values(by='B')#按值排序

	A	B	C	D
2013-01-06	0.892789	-1.848313	0.897434	0.157656
2013-01-05	-0.154101	-0.675057	-0.747016	-0.192082
2013-01-04	-0.507705	-0.409561	-2.596286	0.464993
2013-01-03	0.746757	-0.338140	-0.734583	-2.377116
2013-01-01	-0.828948	0.281765	0.803692	0.030016
2013-01-02	0.418212	1.537528	0.407742	0.625449

选择

df['A']#选择一个列，产生一个“Series”，相当于“df.A”

2013-01-01   -0.828948
2013-01-02    0.418212
2013-01-03    0.746757
2013-01-04   -0.507705
2013-01-05   -0.154101
2013-01-06    0.892789
Freq: D, Name: A, dtype: float64

df[0:3]#通过[ ]选择，对行进行切片

	A	B	C	D
2013-01-01	-0.828948	0.281765	0.803692	0.030016
2013-01-02	0.418212	1.537528	0.407742	0.625449
2013-01-03	0.746757	-0.338140	-0.734583	-2.377116

df['20130102':'20130104']

	A	B	C	D
2013-01-02	0.418212	1.537528	0.407742	0.625449
2013-01-03	0.746757	-0.338140	-0.734583	-2.377116
2013-01-04	-0.507705	-0.409561	-2.596286	0.464993

 df.loc[dates[0]]#通过标签获取一行数据

A   -0.828948
B    0.281765
C    0.803692
D    0.030016
Name: 2013-01-01 00:00:00, dtype: float64

 df.loc['20130102':'20130104',['A','B']]

	A	B
2013-01-02	0.418212	1.537528
2013-01-03	0.746757	-0.338140
2013-01-04	-0.507705	-0.409561

 df.loc['20130102',['A','B']]#减小返回对象的大小

A    0.418212
B    1.537528
Name: 2013-01-02 00:00:00, dtype: float64

 df.at[dates[0],'A']#获取标量值：

-0.8289476073976824

 df.at[dates[0],'A']#快速访问标量

-0.8289476073976824

df.iloc[3]#通过传递的整数的位置选择

A   -0.507705
B   -0.409561
C   -2.596286
D    0.464993
Name: 2013-01-04 00:00:00, dtype: float64

df.iloc[3:5,0:2]#通过整数切片

	A	B
2013-01-04	-0.507705	-0.409561
2013-01-05	-0.154101	-0.675057

df.iloc[[1,2,4],[0,2]]#通过传递整数的列表按位置切片

	A	C
2013-01-02	0.418212	0.407742
2013-01-03	0.746757	-0.734583
2013-01-05	-0.154101	-0.747016

 df.iloc[1,1]#获取具体值

1.5375282822642125

布尔索引

df[df.A > 0]

	A	B	C	D
2013-01-02	0.418212	1.537528	0.407742	0.625449
2013-01-03	0.746757	-0.338140	-0.734583	-2.377116
2013-01-06	0.892789	-1.848313	0.897434	0.157656

df[df > 0]

	A	B	C	D
2013-01-01	NaN	0.281765	0.803692	0.030016
2013-01-02	0.418212	1.537528	0.407742	0.625449
2013-01-03	0.746757	NaN	NaN	NaN
2013-01-04	NaN	NaN	NaN	0.464993
2013-01-05	NaN	NaN	NaN	NaN
2013-01-06	0.892789	NaN	0.897434	0.157656

赋值

s1 = pd.Series([1,2,3,4,5,6], index=pd.date_range('20130102', periods=6))
#添加新列将自动根据索引对齐数据
s1

2013-01-02    1
2013-01-03    2
2013-01-04    3
2013-01-05    4
2013-01-06    5
2013-01-07    6
Freq: D, dtype: int64

df['F'] = s1
df.at[dates[0],'A'] = 0#通过标签赋值
df.iat[0,1] = 0#通过位置赋值
df.loc[:,'D'] = np.array([5] * len(df))
df

	A	B	C	D	F
2013-01-01	0.000000	0.000000	0.803692	5	NaN
2013-01-02	0.418212	1.537528	0.407742	5	1.0
2013-01-03	0.746757	-0.338140	-0.734583	5	2.0
2013-01-04	-0.507705	-0.409561	-2.596286	5	3.0
2013-01-05	-0.154101	-0.675057	-0.747016	5	4.0
2013-01-06	0.892789	-1.848313	0.897434	5	5.0

df2 = df.copy()
df2[df2 > 0] = -df2
df2

	A	B	C	D	F
2013-01-01	0.000000	0.000000	-0.803692	-5	NaN
2013-01-02	-0.418212	-1.537528	-0.407742	-5	-1.0
2013-01-03	-0.746757	-0.338140	-0.734583	-5	-2.0
2013-01-04	-0.507705	-0.409561	-2.596286	-5	-3.0
2013-01-05	-0.154101	-0.675057	-0.747016	-5	-4.0
2013-01-06	-0.892789	-1.848313	-0.897434	-5	-5.0

插入

df.loc['new']=[1,2,3,4,5]
df

	A	B	C	D	F
2013-01-01 00:00:00	0.000000	0.000000	0.803692	5	NaN
2013-01-02 00:00:00	0.418212	1.537528	0.407742	5	1.0
2013-01-03 00:00:00	0.746757	-0.338140	-0.734583	5	2.0
2013-01-04 00:00:00	-0.507705	-0.409561	-2.596286	5	3.0
2013-01-05 00:00:00	-0.154101	-0.675057	-0.747016	5	4.0
2013-01-06 00:00:00	0.892789	-1.848313	0.897434	5	5.0
new	1.000000	2.000000	3.000000	4	5.0

df3=pd.DataFrame([6,6,6,6,6]).T

# 修改df4的column和df3的一致
df3.columns = df.columns
# 把两个dataframe合并，需要设置 ignore_index=True
df_new = pd.concat([df,df3],ignore_index=True)
df_new

	A	B	C	D	F
0	0.000000	0.000000	0.803692	5	NaN
1	0.418212	1.537528	0.407742	5	1.0
2	0.746757	-0.338140	-0.734583	5	2.0
3	-0.507705	-0.409561	-2.596286	5	3.0
4	-0.154101	-0.675057	-0.747016	5	4.0
5	0.892789	-1.848313	0.897434	5	5.0
6	1.000000	2.000000	3.000000	4	5.0
7	6.000000	6.000000	6.000000	6	6.0

统计

df.mean()#平均值

A    0.342279
B    0.038065
C    0.147283
D    4.857143
F    3.333333
dtype: float64

df.mean(1)

2013-01-01 00:00:00    1.450923
2013-01-02 00:00:00    1.672696
2013-01-03 00:00:00    1.334807
2013-01-04 00:00:00    0.897290
2013-01-05 00:00:00    1.484765
2013-01-06 00:00:00    1.988382
new                    3.000000
dtype: float64

df.sum()

A     2.395953
B     0.266457
C     1.030982
D    34.000000
F    20.000000
dtype: float64

df.sum(1)

2013-01-01 00:00:00     5.803692
2013-01-02 00:00:00     8.363482
2013-01-03 00:00:00     6.674034
2013-01-04 00:00:00     4.486448
2013-01-05 00:00:00     7.423826
2013-01-06 00:00:00     9.941910
new                    15.000000
dtype: float64

df.var()#f方差

A    0.331841
B    1.751315
C    3.050677
D    0.142857
F    2.666667
dtype: float64

df.std()#标准差

df.corr()#相关系数

df.cov()#协方差

df.describe()#基本情况

统计作图

import matplotlib.pyplot as plt #导入作图库
plt.rcParams['font.sans-serif'] = ['SimHei']#用来正常显示中文标签
plt.rcParams['axes.unicode_minus'] = False#正常显示负号
plt.figure(figsize = (7,5))#创建作图区域,指定比例

plt.plot(x, y, s)

这是 Matplotlib通用的绘图方式,绘制y对于x(即以x为横轴的二维图形),字符串参量S指定绘制时图形的类型、样式和颜色,常用的选项有:'b’为蓝色、'r’为红色、'g’为绿色、‘o’为圆圈、’+‘为加号标记、’-‘为实线、’–'为虚线。当x、y均为实数同维向量时,则描出点(x(i),y(i)),然后用直线依次相连。

plt.plot(kind=box)

这里使用的是 DataFrame或 Series对象内置的方法作图,默认以 Index为横坐标,每列数据为纵坐标自动作图,通过kind参数指定作图类型,支持line(线)、bar(条形)barh、hist(直方图)、box(箱线图)、kde(密度图)和area、pie(饼图)等,同时也能够接受 plt.plot()中接受的参数。因此,如果数据已经被加载为 Pandas中的对象,那么以这种方式作图是比较简洁的。

x=np.linspace(0,2*np.pi,50) #x坐标
y=np.sin(x)#
plt.plot(x,y,'bp--')
plt.show()

plt.pie(size)

使用Matplotlib绘制饼图,其中size是一个列表,记录各个扇形的比例。pie有丰富的参数.

import matplotlib.pyplot as plt
# The slices will be ordered and plotted counter-clockwise.
labels= 'Frogs', 'Hogs','Dogs', 'Logs' #定义标签
sizes= [15, 30, 45, 10] #每一块的比例
colors=['yellowgreen', 'gold', 'lightskyblue', 'lightcoral'] #每一块的颜色
explode= (0, 0.1, 0, 0) #突出显示,这里仅仅突出显示第二块(即,Hogs' )
plt.pie (sizes, explode=explode, labels=labels, colors=colors, autopct='%1.1f%%',
shadow=True, startangle=90)
plt.axis ('equal') #显示为圆(避免比例压缩为椭圆)
plt.show()