1-1 初识pandas
创建序列
s = pd.Series([1, 3, 6, np.nan, 44, 1])
print(s)
"""
0 1.0
1 3.0
2 6.0
3 NaN
4 44.0
5 1.0
dtype: float64
"""
创建日期型数据，并增长六个日期
date = pd.date_range('2016-01-01', periods = 6)
print(date)
"""
DatetimeIndex(['2016-01-01', '2016-01-02', '2016-01-03', '2016-01-04',
'2016-01-05', '2016-01-06'],
dtype='datetime64[ns]', freq='D')
"""
创建有索引和字符串的数据
df = pd.DataFrame(np.random.rand(6,4), index=date, columns=['a', 'b', 'c', 'd'])
print(df)
"""
a b c d
2016-01-01 0.113951 0.583000 0.167336 0.917897
2016-01-02 0.632843 0.950597 0.280311 0.946806
2016-01-03 0.367501 0.313236 0.475095 0.889570
2016-01-04 0.653676 0.444720 0.091550 0.272699
2016-01-05 0.448919 0.328602 0.644945 0.196358
2016-01-06 0.656723 0.355628 0.886951 0.688788
"""
不加索引的矩阵
df1 = pd.DataFrame(np.arange(12).reshape((3,4)))
print(df1)
# 创建决定各个数值的矩阵 categorical 明确的，确定的
df2 = pd.DataFrame({'A':1,
'B':pd.Timestamp('20120202'),
'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'
})
print(df2)
"""
A B C D E F
0 1 2012-02-02 1.0 3 test foo
1 1 2012-02-02 1.0 3 train foo
2 1 2012-02-02 1.0 3 test foo
3 1 2012-02-02 1.0 3 train foo
"""
# df2 的dtype
print(df2.dtypes)
"""
A int64
B datetime64[ns]
C float32
D int32
E category
F object
dtype: object
"""
# 列标
print(df2.columns)
"Index(['A', 'B', 'C', 'D', 'E', 'F'], dtype='object')"
# 值
print(df2.values)
"""
[[1 Timestamp('2012-02-02 00:00:00') 1.0 3 'test' 'foo']
[1 Timestamp('2012-02-02 00:00:00') 1.0 3 'train' 'foo']
[1 Timestamp('2012-02-02 00:00:00') 1.0 3 'test' 'foo']
[1 Timestamp('2012-02-02 00:00:00') 1.0 3 'train' 'foo']]
"""
# 描述
print(df2.describe())
"""
A C D
count 4.0 4.0 4.0
mean 1.0 1.0 3.0
std 0.0 0.0 0.0
min 1.0 1.0 3.0
25% 1.0 1.0 3.0
50% 1.0 1.0 3.0
75% 1.0 1.0 3.0
max 1.0 1.0 3.0
"""
# 转置
print(df2.T)
# 按列或者行排序
print(df2.sort_index(axis=1, ascending=False))
"""
F E D C B A
0 foo test 3 1.0 2012-02-02 1
1 foo train 3 1.0 2012-02-02 1
2 foo test 3 1.0 2012-02-02 1
3 foo train 3 1.0 2012-02-02 1
"""
# 输出部分索引的值并排序
print(df2.sort_values(by='E'))
"""
A B C D E F
0 1 2012-02-02 1.0 3 test foo
2 1 2012-02-02 1.0 3 test foo
1 1 2012-02-02 1.0 3 train foo
3 1 2012-02-02 1.0 3 train foonn
"""
1-2 数据处理
# 先创建数据备用
date = pd.date_range('20130101', periods=6)
df = pd.DataFrame(np.arange(24).reshape((6,4)), index=date, columns=['A', 'B', 'C', 'D'])
print(df)
# 多种方式打印
print(df['A'])
print(df.A)
# 打印前三行
print(df[0:3])
# print('\n'*1)
# 打印后三行
print(df['2013-01-04':'2013-01-06'])
# select by label:loc
print(df.loc['2013-01-06'])
"""
A 20
B 21
C 22
D 23
Name: 2013-01-06 00:00:00, dtype: int32
"""
print(df.loc[:,['A', 'B']])
"""
A B
2013-01-01 0 1
2013-01-02 4 5
2013-01-03 8 9
2013-01-04 12 13
2013-01-05 16 17
2013-01-06 20 21
"""
print(df.loc['2013-01-03', ['A', 'B']])
"""
A 8
B 9
Name: 2013-01-03 00:00:00, dtype: int32
"""
# select by of position：iloc
print(df.iloc[[1, 3, 5], 1:3])
"""
B C
2013-01-02 5 6
2013-01-04 13 14
2013-01-06 21 22
"""
# select mixed: ix
# print(df.ix[:3, ['A', 'B', 'C']])
# boolean indexing
print(df)
print(df<8)
print(df[df.A<8])
"""
A B C D
2013-01-01 True True True True
2013-01-02 True True True True
2013-01-03 False False False False
2013-01-04 False False False False
2013-01-05 False False False False
2013-01-06 False False False False
A B C D
2013-01-01 0 1 2 3
2013-01-02 4 5 6 7
"""
1-3 设置数据值
date = pd.date_range('20100101', periods=6)
df = pd.DataFrame(np.arange(24).reshape((6, 4)), index=date, columns=['A', 'B', 'C', 'D'])
print(df)
# set values by position: iloc
df.iloc[2,2] = 111
print(df)
# set values by label: loc
df.loc['20100103', 'B'] = 999
print(df)
# 设置>大于某个数的值 改变全部值
# df[df.A>0] = 0
# print(df)
"""
A B C D
2010-01-01 0 1 2 3
2010-01-02 0 0 0 0
2010-01-03 0 0 0 0
2010-01-04 0 0 0 0
2010-01-05 0 0 0 0
2010-01-06 0 0 0 0
"""
# 改变一部分的值
# df.A[df.A>3] = 0
# print(df)
"""
A B C D
2010-01-01 0 1 2 3
2010-01-02 0 5 6 7
2010-01-03 0 999 111 11
2010-01-04 0 13 14 15
2010-01-05 0 17 18 19
2010-01-06 0 21 22 23
"""
# 增加值
df['F'] = np.nan
print(df)
"""
A B C D F
2010-01-01 0 1 2 3 NaN
2010-01-02 4 5 6 7 NaN
2010-01-03 8 999 111 11 NaN
2010-01-04 12 13 14 15 NaN
2010-01-05 16 17 18 19 NaN
2010-01-06 20 21 22 23 NaN
"""
df['E'] = pd.Series([1,2,3,4,5,6], index=pd.date_range('20100101', periods=6))
print(df)
A B C D F E
2010-01-01 0 1 2 3 NaN 1
2010-01-02 4 5 6 7 NaN 2
2010-01-03 8 999 111 11 NaN 3
2010-01-04 12 13 14 15 NaN 4
2010-01-05 16 17 18 19 NaN 5
2010-01-06 20 21 22 23 NaN 6
1-4 处理丢失的数据
import numpy as np
import pandas as pd
# 原始数据
date = pd.date_range('20100101', periods=6)
df = pd.DataFrame(np.arange(24).reshape((6, 4)), index=date, columns=['A', 'B', 'C', 'D'])
print(df)
df.iloc[0, 1] = np.nan
df.iloc[1, 2] = np.nan
print(df)
# 处理数据-丢掉有nan的整行数据 any-只要行或列中年有nan就算，all-要全部都为nan才算
print(df.dropna(axis=0, how='any')) # axis = 0， 竖直方向对行进行操作
print(df.dropna(axis=1, how='any')) # axis = 1 水平方向对列进行操作
"""
A B C D
2010-01-03 8 9.0 10.0 11
2010-01-04 12 13.0 14.0 15
2010-01-05 16 17.0 18.0 19
2010-01-06 20 21.0 22.0 23
A D
2010-01-01 0 3
2010-01-02 4 7
2010-01-03 8 11
2010-01-04 12 15
2010-01-05 16 19
2010-01-06 20 23
"""
# 将nan的位置填入其他值，如0
print(df.fillna(value=0))
"""
A B C D
2010-01-01 0 0.0 2.0 3
2010-01-02 4 5.0 0.0 7
2010-01-03 8 9.0 10.0 11
2010-01-04 12 13.0 14.0 15
2010-01-05 16 17.0 18.0 19
2010-01-06 20 21.0 22.0 23
"""
# 打印True或者False的值
print(df.isnull())
"""
A B C D
2010-01-01 False True False False
2010-01-02 False False True False
2010-01-03 False False False False
2010-01-04 False False False False
2010-01-05 False False False False
2010-01-06 False False False False
"""
# 打印至少一个是丢失的数据的判断情况
print(np.any(df.isnull())) # True
1-5 导入与导出数据
# 将csv文件读入，该文件可以再=在任意目录，将其读入就可以
data = pd.read_csv('C:/Users/liyuelong/Desktop/student.csv')
print(data)
# 转成其他模式，目录地址要相同。
data.to_pickle('C:/Users/liyuelong/Desktop/student.pickle')