数据类型
import numpy as np
import random
t1 = np.array([1, 2, 3])
print(t1, type(t1)) # [1 2 3] <class 'numpy.ndarray'>
t2 = np.array(range(5))
print(t2, type(t2)) # [0 1 2 3 4] <class 'numpy.ndarray'>
t3 = np.arange(5) # arange用法参数类range
print(t3, type(t3)) # [0 1 2 3 4] <class 'numpy.ndarray'>
print(t3.dtype) # int64
# numpy数据类型
t4 = np.array(range(1, 4), dtype=float) # dtype指定类型
print(t4, t4.dtype) # [1. 2. 3.] float64
# numpy bool类型
t5 = np.array([1, 0, 0, 1, 0], dtype=bool)
print(t5, t5.dtype) # [ True False False True False] bool
# 调整数据类型
t6 = t5.astype('int8')
print(t6, t6.dtype) # [1 0 0 1 0] int8
# numpy小数
t7 = np.array([random.random() for i in range(10)])
print(t7, t7.dtype) # [0.84702583 0.916558 0.44216734 0.53020263 0.44274757 0.0559538 0.53722744 0.04059448 0.70912489 0.94199106] float64
t8 = np.round(t7, 2)
print(t8, t8.dtype) # [0.85 0.92 0.44 0.53 0.44 0.06 0.54 0.04 0.71 0.94] float64
t9 = '%.2f'%random.random() # %:占位符, 2:保留二位小数, f:浮点型
print(t9, type(t9)) # 0.65 <class 'str'>
In [1]: import numpy as np
In [2]: t1 = np.arange(12)
In [3]: t1
Out[3]: array([ 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11])
In [4]: t1.shape # 查看数组形状
Out[4]: (12,) # arrary() 参数只有一个列表时, 一维数组，t1.shape元祖一个值,12为列表元素数量
In [5]: t2 = np.array([[1,2,3],[4,5,6]])
In [6]: t2
Out[6]:
array([[1, 2, 3],
[4, 5, 6]])
In [7]: t2.shape
Out[7]: (2, 3) # arrary() 参数只有二个列表时, 二维数组,t1.shape元祖有二个值,(2, 3), 2行数，3列数
In [8]: t3 = np.array([[[1,2,3],[4,5,6]],[[7,8,9],[10,11,12]]])
In [9]: t3
Out[9]:
array([[[ 1, 2, 3],
[ 4, 5, 6]],
[[ 7, 8, 9],
[10, 11, 12]]])
In [10]: t3.shape # t3为三维数组,t3.shape元祖有三个值
Out[10]: (2, 2, 3)
In [11]: t4 = np.arange(12)
In [12]: t4
Out[12]: array([ 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11])
In [13]: t4.reshape((3,4)) # 修改数组形状,变成3行4列
Out[13]:
array([[ 0, 1, 2, 3],
[ 4, 5, 6, 7],
[ 8, 9, 10, 11]])
In [14]: t4.reshape((3,5))
---------------------------------------------------------------------------
ValueError Traceback (most recent call last)
<ipython-input-14-7634752f2bda> in <module>
----> 1 t4.reshape((3,5))
ValueError: cannot reshape array of size 12 into shape (3,5)
In [15]: t5 = np.arange(24).reshape((2,3,4)) # (2,3,4):2块,3行,4列
In [16]: t5
Out[16]:
array([[[ 0, 1, 2, 3],
[ 4, 5, 6, 7],
[ 8, 9, 10, 11]],
[[12, 13, 14, 15],
[16, 17, 18, 19],
[20, 21, 22, 23]]])
In [17]: t5.reshape((4,6))
Out[17]:
array([[ 0, 1, 2, 3, 4, 5],
[ 6, 7, 8, 9, 10, 11],
[12, 13, 14, 15, 16, 17],
[18, 19, 20, 21, 22, 23]])
In [18]: t5
Out[18]:
array([[[ 0, 1, 2, 3],
[ 4, 5, 6, 7],
[ 8, 9, 10, 11]],
[[12, 13, 14, 15],
[16, 17, 18, 19],
[20, 21, 22, 23]]])
In [19]: t5 = t5.reshape((4,6))
In [20]: t5
Out[20]:
array([[ 0, 1, 2, 3, 4, 5],
[ 6, 7, 8, 9, 10, 11],
[12, 13, 14, 15, 16, 17],
[18, 19, 20, 21, 22, 23]])
In [21]: t5.reshape((24,))
Out[21]:
array([ 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16,
17, 18, 19, 20, 21, 22, 23])
In [22]: t5.reshape((24,1))
Out[22]:
array([[ 0],
[ 1],
[ 2],
[ 3],
[ 4],
[ 5],
[ 6],
[ 7],
[ 8],
[ 9],
[10],
[11],
[12],
[13],
[14],
[15],
[16],
[17],
[18],
[19],
[20],
[21],
[22],
[23]])
In [23]: t5.reshape((1,24))
Out[23]:
array([[ 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15,
16, 17, 18, 19, 20, 21, 22, 23]])
In [24]: t6 = t5.reshape((t5.shape[0]*t5.shape[1],)) # t5.shape[0]:t5行数,t5.shape[1]:t5列数
In [25]: t6
Out[25]:
array([ 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16,
17, 18, 19, 20, 21, 22, 23])
In [26]: t5
Out[26]:
array([[ 0, 1, 2, 3, 4, 5],
[ 6, 7, 8, 9, 10, 11],
[12, 13, 14, 15, 16, 17],
[18, 19, 20, 21, 22, 23]])
In [27]: t5.flatten() # t5.flatten()数据转为一维数组
Out[27]:
array([ 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16,
17, 18, 19, 20, 21, 22, 23])
In [28]: t5
Out[28]:
array([[ 0, 1, 2, 3, 4, 5],
[ 6, 7, 8, 9, 10, 11],
[12, 13, 14, 15, 16, 17],
[18, 19, 20, 21, 22, 23]])
In [29]: t5+2
Out[29]:
array([[ 2, 3, 4, 5, 6, 7],
[ 8, 9, 10, 11, 12, 13],
[14, 15, 16, 17, 18, 19],
[20, 21, 22, 23, 24, 25]])
In [30]: t5*2
Out[30]:
array([[ 0, 2, 4, 6, 8, 10],
[12, 14, 16, 18, 20, 22],
[24, 26, 28, 30, 32, 34],
[36, 38, 40, 42, 44, 46]])
In [31]: t5/2
Out[31]:
array([[ 0. , 0.5, 1. , 1.5, 2. , 2.5],
[ 3. , 3.5, 4. , 4.5, 5. , 5.5],
[ 6. , 6.5, 7. , 7.5, 8. , 8.5],
[ 9. , 9.5, 10. , 10.5, 11. , 11.5]])
In [32]: t5/0
/Users/xyp/opt/anaconda3/envs/DataAnalysis/bin/ipython:1: RuntimeWarning: divide by zero encountered in true_divide
#!/Users/xyp/opt/anaconda3/envs/DataAnalysis/bin/python
/Users/xyp/opt/anaconda3/envs/DataAnalysis/bin/ipython:1: RuntimeWarning: invalid value encountered in true_divide
#!/Users/xyp/opt/anaconda3/envs/DataAnalysis/bin/python
Out[32]: # inf +∞
array([[nan, inf, inf, inf, inf, inf],
[inf, inf, inf, inf, inf, inf],
[inf, inf, inf, inf, inf, inf],
[inf, inf, inf, inf, inf, inf]])
In [33]: t5
Out[33]:
array([[ 0, 1, 2, 3, 4, 5],
[ 6, 7, 8, 9, 10, 11],
[12, 13, 14, 15, 16, 17],
[18, 19, 20, 21, 22, 23]])
In [34]: t6 = np.arange(100,124).reshape((4,6))
In [35]: t6
Out[35]:
array([[100, 101, 102, 103, 104, 105],
[106, 107, 108, 109, 110, 111],
[112, 113, 114, 115, 116, 117],
[118, 119, 120, 121, 122, 123]])
In [36]: t5+t6
Out[36]:
array([[100, 102, 104, 106, 108, 110],
[112, 114, 116, 118, 120, 122],
[124, 126, 128, 130, 132, 134],
[136, 138, 140, 142, 144, 146]])
In [37]: t5*t6
Out[37]:
array([[ 0, 101, 204, 309, 416, 525],
[ 636, 749, 864, 981, 1100, 1221],
[1344, 1469, 1596, 1725, 1856, 1989],
[2124, 2261, 2400, 2541, 2684, 2829]])
In [38]: t6/t5
/Users/xyp/opt/anaconda3/envs/DataAnalysis/bin/ipython:1: RuntimeWarning: divide by zero encountered in true_divide
#!/Users/xyp/opt/anaconda3/envs/DataAnalysis/bin/python
Out[38]:
array([[ inf, 101. , 51. , 34.33333333,
26. , 21. ],
[ 17.66666667, 15.28571429, 13.5 , 12.11111111,
11. , 10.09090909],
[ 9.33333333, 8.69230769, 8.14285714, 7.66666667,
7.25 , 6.88235294],
[ 6.55555556, 6.26315789, 6. , 5.76190476,
5.54545455, 5.34782609]])
In [39]: t7 = np.arange(0,6)
In [40]: t7
Out[40]: array([0, 1, 2, 3, 4, 5])
In [41]: t5
Out[41]:
array([[ 0, 1, 2, 3, 4, 5],
[ 6, 7, 8, 9, 10, 11],
[12, 13, 14, 15, 16, 17],
[18, 19, 20, 21, 22, 23]])
In [42]: t5-t7
Out[42]:
array([[ 0, 0, 0, 0, 0, 0],
[ 6, 6, 6, 6, 6, 6],
[12, 12, 12, 12, 12, 12],
[18, 18, 18, 18, 18, 18]])
In [43]: t8 = np.arange(4).reshape((4,1))
In [44]: t8
Out[44]:
array([[0],
[1],
[2],
[3]])
In [45]: t5-t8
Out[45]:
array([[ 0, 1, 2, 3, 4, 5],
[ 5, 6, 7, 8, 9, 10],
[10, 11, 12, 13, 14, 15],
[15, 16, 17, 18, 19, 20]])
In [46]: t9 = np.arange(10)
In [47]: t9
Out[47]: array([0, 1, 2, 3, 4, 5, 6, 7, 8, 9])
In [48]: t5
Out[48]:
array([[ 0, 1, 2, 3, 4, 5],
[ 6, 7, 8, 9, 10, 11],
[12, 13, 14, 15, 16, 17],
[18, 19, 20, 21, 22, 23]])
In [49]: t5-t9
---------------------------------------------------------------------------
ValueError Traceback (most recent call last)
<ipython-input-49-dcf9c8dd3788> in <module>
----> 1 t5-t9
ValueError: operands could not be broadcast together with shapes (4,6) (10,)
文件读取
import numpy as np
us_file_path = './youtube_video_data/US_video_data_numbers.csv'
# delimiter读取csv用,分割;unpack默认False,按行读取数据,unpack=True按列读取数据,转置;dtype数据类型,默认科学计数方式;skiprows跳过前X行;usecols读取指定列,索引,元祖类型
t1 = np.loadtxt(us_file_path, delimiter=',', dtype=np.int)
print(t1, '\n', '*'*100)
t2 = np.loadtxt(us_file_path, delimiter=',', dtype=np.int, unpack=True)
print(t2)
[[4394029 320053 5931 46245]
[7860119 185853 26679 0]
[5845909 576597 39774 170708]
...
[ 142463 4231 148 279]
[2162240 41032 1384 4737]
[ 515000 34727 195 4722]]
****************************************************************************************************
[[4394029 7860119 5845909 ... 142463 2162240 515000]
[ 320053 185853 576597 ... 4231 41032 34727]
[ 5931 26679 39774 ... 148 1384 195]
[ 46245 0 170708 ... 279 4737 4722]]
转置三种方法和简单运算
# 转置三种方法
In [2]: import numpy as np
In [3]: t2 = np.arange(24).reshape((4,6))
In [4]: t2
Out[4]:
array([[ 0, 1, 2, 3, 4, 5],
[ 6, 7, 8, 9, 10, 11],
[12, 13, 14, 15, 16, 17],
[18, 19, 20, 21, 22, 23]])
In [6]: t2.transpose() # 转置
Out[6]:
array([[ 0, 6, 12, 18],
[ 1, 7, 13, 19],
[ 2, 8, 14, 20],
[ 3, 9, 15, 21],
[ 4, 10, 16, 22],
[ 5, 11, 17, 23]])
In [7]: t2.T # 转置
Out[7]:
array([[ 0, 6, 12, 18],
[ 1, 7, 13, 19],
[ 2, 8, 14, 20],
[ 3, 9, 15, 21],
[ 4, 10, 16, 22],
[ 5, 11, 17, 23]])
In [8]: t2.swapaxes(1,0) # 交换轴
Out[8]:
array([[ 0, 6, 12, 18],
[ 1, 7, 13, 19],
[ 2, 8, 14, 20],
[ 3, 9, 15, 21],
[ 4, 10, 16, 22],
[ 5, 11, 17, 23]])
In [9]: t2
Out[9]:
array([[ 0, 1, 2, 3, 4, 5],
[ 6, 7, 8, 9, 10, 11],
[12, 13, 14, 15, 16, 17],
[18, 19, 20, 21, 22, 23]])
In [10]: t2<10
Out[10]:
array([[ True, True, True, True, True, True],
[ True, True, True, True, False, False],
[False, False, False, False, False, False],
[False, False, False, False, False, False]])
In [11]: t2[t2<10]=3
In [12]: t2
Out[12]:
array([[ 3, 3, 3, 3, 3, 3],
[ 3, 3, 3, 3, 10, 11],
[12, 13, 14, 15, 16, 17],
[18, 19, 20, 21, 22, 23]])
In [13]: t2[t2>20]
Out[13]: array([21, 22, 23])
In [16]: t2
Out[16]:
array([[ 3, 3, 3, 3, 3, 3],
[ 3, 3, 3, 3, 10, 11],
[12, 13, 14, 15, 16, 17],
[18, 19, 20, 21, 22, 23]])
In [17]: np.where(t2<=3,100,300) # numpy三元运算符,t2 = 100 if t2<=3 else 300
Out[17]:
array([[100, 100, 100, 100, 100, 100],
[100, 100, 100, 100, 300, 300],
[300, 300, 300, 300, 300, 300],
[300, 300, 300, 300, 300, 300]])
In [19]: t = np.arange(20)
In [20]: t
Out[20]:
array([ 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16,
17, 18, 19])
In [21]: t.clip(10,18) # clip(10,18),小于10的替换成10,大于18的替换成18
Out[21]:
array([10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 11, 12, 13, 14, 15, 16,
17, 18, 18])
In [22]: t[2]=np.nan
---------------------------------------------------------------------------
ValueError Traceback (most recent call last)
<ipython-input-22-1aa5d7dd59fe> in <module>
----> 1 t[2]=np.nan
ValueError: cannot convert float NaN to integer
In [23]: t=t.astype(float)
In [24]: t[2]=np.nan
In [25]: t
Out[25]:
array([ 0., 1., nan, 3., 4., 5., 6., 7., 8., 9., 10., 11., 12.,
13., 14., 15., 16., 17., 18., 19.])
索引和切片
import numpy as np
us_file_path = './youtube_video_data/US_video_data_numbers.csv'
uk_file_path = './youtube_video_data/GB_video_data_numbers.csv'
# delimiter读取csv用,分割;unpack默认False,按行读取数据,unpack=True按列读取数据,转置;dtype数据类型,默认科学计数方式;skiprows跳过前X行;usecols读取指定列,索引,元祖类型
t1 = np.loadtxt(us_file_path, delimiter=',', dtype=np.int)
print(t1, '\n', '*'*100)
t2 = np.loadtxt(us_file_path, delimiter=',', dtype=np.int, unpack=True)
# print(t2)
# 取行
# print(t1[2])
# 取多行
# print(t1[2:])
# 取指定行,2,8,10为索引
print(t1[[2, 8, 10]], '\n', '*'*100)
# 取连续列和列,[1, :]、[2, :]、[[2, 3, 10], :]、[2, [0, 2]] 逗号前为指定行，逗号后为指定列
# print(t1[1, :])
# print(t1[2, :])
# print(t1[[2, 3, 10], :])
# print(t1[2, [0, 2]])
print(t1[2:5, 1:4], '\n', '*'*100)
# 取指定行和列,[[0, 2, 3], [0, 1, 3]]不是取索引为0,2,3的行和0,1,3的列,取的是t1坐标(0,2),(2,1),(3,3)的数据
print(t1[[0, 2, 3], [0, 1, 3]])
[[4394029 320053 5931 46245]
[7860119 185853 26679 0]
[5845909 576597 39774 170708]
...
[ 142463 4231 148 279]
[2162240 41032 1384 4737]
[ 515000 34727 195 4722]]
****************************************************************************************************
[[5845909 576597 39774 170708]
[1338533 69687 678 5643]
[ 859289 34485 726 1914]]
****************************************************************************************************
[[576597 39774 170708]
[ 24975 4542 12829]
[ 96666 568 6666]]
****************************************************************************************************
[4394029 576597 12829]
数组的拼接
In [31]: t1
Out[31]:
array([[0, 1, 2, 3],
[4, 5, 6, 7]])
In [32]: t2
Out[32]:
array([[ 8, 9, 10, 11],
[12, 13, 14, 15]])
In [33]: np.vstack((t1,t2)) # 竖直拼接
Out[33]:
array([[ 0, 1, 2, 3],
[ 4, 5, 6, 7],
[ 8, 9, 10, 11],
[12, 13, 14, 15]])
In [34]: np.hstack((t1,t2)) # 水平拼接
Out[34]:
array([[ 0, 1, 2, 3, 8, 9, 10, 11],
[ 4, 5, 6, 7, 12, 13, 14, 15]])
数组的行列交换
In [39]: t = np.arange(12,24).reshape(3,4)
In [40]: t
Out[40]:
array([[12, 13, 14, 15],
[16, 17, 18, 19],
[20, 21, 22, 23]])
In [41]: t[[1,2],:] = t[[2,1],:] # 行交换
In [42]: t
Out[42]:
array([[12, 13, 14, 15],
[20, 21, 22, 23],
[16, 17, 18, 19]])
In [43]: t[:,[0,2]] = t[:,[2,0]] # 列交换
In [44]: t
Out[44]:
array([[14, 13, 12, 15],
[22, 21, 20, 23],
[18, 17, 16, 19]])
numpy中nan和常用方法
In [39]: np.nan == np.nan
Out[39]: False
In [40]: np.nan != np.nan
Out[40]: True
In [49]: t
Out[49]:
array([[0., 1., 2., 3., 4.],
[5., 6., 7., 8., 9.]])
In [50]: t[:,0] = 0
In [51]: t
Out[51]:
array([[0., 1., 2., 3., 4.],
[0., 6., 7., 8., 9.]])
In [52]: np.count_nonzero(t) # 判断t数组中非0个数
Out[52]: 8
In [56]: t[:,0] = np.nan
In [57]: t
Out[57]:
array([[nan, 1., 2., 3., 4.],
[nan, 6., 7., 8., 9.]])
In [58]: t != t # 当t!=t时,数组中为nan时为True
Out[58]:
array([[ True, False, False, False, False],
[ True, False, False, False, False]])
In [59]: np.count_nonzero(t!=t) # 数组t中为nan的个数
Out[59]: 2
In [60]: np.isnan(t) # 数组中为nan时为True
Out[60]:
array([[ True, False, False, False, False],
[ True, False, False, False, False]])
In [61]: np.count_nonzero(np.isnan(t)) # 数组t中为nan的个数
Out[61]: 2
In [62]: np.sum(t) # nan和任何值的计算都为nan
Out[62]: nan
In [63]: np.sum(t,axis=0) # sum(t,axis=0) 列相加结果
Out[63]: array([nan, 7., 9., 11., 13.])
In [64]: np.sum(t,axis=1) # sum(t,axis=1) 行相加结果
Out[64]: array([nan, nan])
numpy中常用统计方法
In [75]: t
Out[75]:
array([[0, 1, 2, 3, 4],
[5, 6, 7, 8, 9]])
In [76]: t.sum(axis=0)
Out[76]: array([ 5, 7, 9, 11, 13])
In [77]: t.sum(axis=1)
Out[77]: array([10, 35])
In [78]: t.mean(axis=0)
Out[78]: array([2.5, 3.5, 4.5, 5.5, 6.5])
In [79]: t.mean(axis=1) # 均值,当数组中有nan时,剩余非nan元素的均值替换nan
Out[79]: array([2., 7.])
In [80]: np.median(t) # 中值
Out[80]: 4.5
In [81]: np.median(t,axis=0)
Out[81]: array([2.5, 3.5, 4.5, 5.5, 6.5])
In [82]: np.median(t,axis=1)
Out[82]: array([2., 7.])
In [83]: t.max()
Out[83]: 9
In [84]: t.max(axis=0)
Out[84]: array([5, 6, 7, 8, 9])
In [85]: t.max(axis=1)
Out[85]: array([4, 9])
In [86]: t.min(axis=1)
Out[86]: array([0, 5])
In [87]: np.ptp(t,axis=0) # 极值，最大值和最小值差
Out[87]: array([5, 5, 5, 5, 5])
In [88]: np.ptp(t,axis=1)
Out[88]: array([4, 4])
In [92]: t
Out[92]:
array([[0, 1, 2, 3, 4],
[5, 6, 7, 8, 9]])
In [93]: t.std() # 标准差
Out[93]: 2.8722813232690143
In [94]: t.std(axis=0)
Out[94]: array([2.5, 2.5, 2.5, 2.5, 2.5])
In [95]: t.std(axis=1)
Out[95]: array([1.41421356, 1.41421356])
import numpy as np
us_file_path = './youtube_video_data/US_video_data_numbers.csv'
uk_file_path = './youtube_video_data/GB_video_data_numbers.csv'
# 加载国家数据
us_data = np.loadtxt(us_file_path, delimiter=',', dtype='int')
uk_data = np.loadtxt(uk_file_path, delimiter=',', dtype='int')
# 添加国家信息
# 构造全为0数据
zeros_data = np.zeros((us_data.shape[0], 1)).astype(int)
ones_data = np.ones((uk_data.shape[0], 1)).astype(int)
# 分别添加一列全为0,1的数组
us_data = np.hstack((us_data, zeros_data))
uk_data = np.hstack((uk_data, ones_data))
# 拼接两组数据,最后一列全为0的代表us,为1的代表uk
final_data = np.vstack((us_data, uk_data))
print(final_data)
# 把二个国家数据拼接一起研究分析数据
import numpy as np
us_file_path = './youtube_video_data/US_video_data_numbers.csv'
uk_file_path = './youtube_video_data/GB_video_data_numbers.csv'
# 加载国家数据
us_data = np.loadtxt(us_file_path, delimiter=',', dtype='int')
uk_data = np.loadtxt(uk_file_path, delimiter=',', dtype='int')
# 添加国家信息
# 构造全为0数据
zeros_data = np.zeros((us_data.shape[0], 1)).astype(int)
ones_data = np.ones((uk_data.shape[0], 1)).astype(int)
# 分别添加一列全为0,1的数组
us_data = np.hstack((us_data, zeros_data))
uk_data = np.hstack((uk_data, ones_data))
# 拼接两组数据,最后一列全为0的代表us,为1的代表uk
final_data = np.vstack((us_data, uk_data))
print(final_data)
把二个国家数据拼接
# numpy更多方法
In [1]: import numpy as np
In [2]: np.ones((3,4)) # 创建全为1的数组
Out[2]:
array([[1., 1., 1., 1.],
[1., 1., 1., 1.],
[1., 1., 1., 1.]])
In [3]: np.zeros((3,4)) # 创建全为0的数组
Out[3]:
array([[0., 0., 0., 0.],
[0., 0., 0., 0.],
[0., 0., 0., 0.]])
In [4]: np.eye(5) # 创建对角线全为1的正方形数组
Out[4]:
array([[1., 0., 0., 0., 0.],
[0., 1., 0., 0., 0.],
[0., 0., 1., 0., 0.],
[0., 0., 0., 1., 0.],
[0., 0., 0., 0., 1.]])
In [5]: t = np.eye(5)
In [6]: np.argmax(t,axis=0) # 获取最大值位置
Out[6]: array([0, 1, 2, 3, 4])
In [7]: t[t==1] = -1
In [8]: t
Out[8]:
array([[-1., 0., 0., 0., 0.],
[ 0., -1., 0., 0., 0.],
[ 0., 0., -1., 0., 0.],
[ 0., 0., 0., -1., 0.],
[ 0., 0., 0., 0., -1.]])
In [9]: np.argmin(t,axis=1) # 获取最小值位置
Out[9]: array([0, 1, 2, 3, 4])
numpy随机方法
In [12]: np.random.rand(2,3) # 创建二维三列的均匀分布范围0～1浮点型数组
Out[12]:
array([[0.06364283, 0.91082238, 0.78795567],
[0.0627046 , 0.33476692, 0.5778516 ]])
In [13]: np.random.randn(2,3) # 创建二维三列的标准正态分布随机数,平均数为0标准差为1浮点型数组
Out[13]:
array([[ 0.06391798, 0.19011529, -0.17431257],
[-0.45543116, -0.02290774, 0.11979098]])
In [14]: np.random.randint(0,100,(2,3)) # 创建二维三列最低为0最高为99的随机整数
Out[14]:
array([[13, 98, 78],
[36, 59, 97]])
In [20]: np.random.uniform(0,5,(2,3)) # 创建二维三列最低为0最高为99的随机浮点型小数
Out[20]:
array([[4.95142868, 1.39926247, 4.21451073],
[2.25316875, 2.67873448, 2.84466319]])
np.random.seed(0) # 随机数种子，这样每次生成相同的随机数，参数为给定的种子值
t = np.random.randint(0, 10, (2, 3))
print(t)
