索引与切片

索引与切片.png

总结一下Numpy中索引与切片的常用操作，思维导图可以帮助自己快速梳理回顾知识点。个人觉得Numpy作为一个工具，没必要花太多精力去熟悉每个API，遇到没见过的API查看官方文档学习即可。

一维数组

a = np.arange(10)
a

array([0, 1, 2, 3, 4, 5, 6, 7, 8, 9])

索引

a[0] = 1
print(a, a[1])

[1 1 2 3 4 5 6 7 8 9] 1

切片

a[0:2] = 2
a

array([2, 2, 2, 3, 4, 5, 6, 7, 8, 9])

# 拷贝数组
b = a[:].copy()
b[0] = 1
print(a == b)

[False  True  True  True  True  True  True  True  True  True]

多维数组

a = np.arange(12).reshape(3,-1)
a

array([[ 0,  1,  2,  3],
       [ 4,  5,  6,  7],
       [ 8,  9, 10, 11]])

索引

# ndim从原来的2-->1
print(a[0])
print(a[0, 0])  # 等价于a[0][0]
print('-------')
print(a.ndim)
print(a[0].ndim)
print(a[0, 0].ndim)

[0 1 2 3]
0
-------
2
1
0

dot 索引

b = np.random.randint(1, 100, [2, 2, 3])
print(b)
print(b[1, ...])
print(b[..., 2])

[[[35 86 56]
  [18  9 50]]

 [[29 94  2]
  [48 26 26]]]
[[29 94  2]
 [48 26 26]]
[[56 50]
 [ 2 26]]

切片

# 切片不改变数组的维度
print(a[:2, 1:])

[[1 2 3]
 [5 6 7]]

# 索引 & 切片
# 得到降低一个维度的切片
print(a[1,:2])

[4 5]

布尔索引

data = np.random.randn(7, 4)
names = np.array(['Bob', 'Joe', 'Will', 'Bob', 'Will', 'Joe', 'Joe'])

data

array([[-0.7508834 , -0.54532566, -0.76911561, -1.62799584],
       [ 1.83550152, -0.90720255,  0.5720965 ,  0.23752405],
       [ 1.02863757, -0.03463182, -0.26193802,  1.60239582],
       [ 0.42237715, -1.4393293 ,  1.57730021, -0.90482475],
       [-0.09214181,  0.17379535,  0.40528223, -2.7919602 ],
       [-0.33572487, -0.74226905,  1.16949952,  0.25240986],
       [-1.58267947, -0.53052068,  0.20950701, -1.9841287 ]])

names

array(['Bob', 'Joe', 'Will', 'Bob', 'Will', 'Joe', 'Joe'], dtype='<U4')

names == 'Bob'

array([ True, False, False,  True, False, False, False])

data[names=='Bob']

array([[-0.7508834 , -0.54532566, -0.76911561, -1.62799584],
       [ 0.42237715, -1.4393293 ,  1.57730021, -0.90482475]])

data[names=='Bob', 2:]

array([[-0.76911561, -1.62799584],
       [ 1.57730021, -0.90482475]])

神奇索引

a = np.arange(12).reshape(4, -1)
a

array([[ 0,  1,  2],
       [ 3,  4,  5],
       [ 6,  7,  8],
       [ 9, 10, 11]])

a[[3, 2, 0, 1]]

array([[ 9, 10, 11],
       [ 6,  7,  8],
       [ 0,  1,  2],
       [ 3,  4,  5]])

a[[3, 2, 1], [2, 0, 1]]

array([11,  6,  4])

a[0, [1, 2]]

array([1, 2])

a[:2, [0, 2]]

array([[0, 2],
       [3, 5]])

a[:2, [True, False, True]]

array([[0, 2],
       [3, 5]])