numpy.meshgrid(x,y)
[X,Y] = meshgrid(x,y)
生成size(y)Xsize(x)大小的矩阵X和Y。它相当于x从一行重复增加到size(y)行，把y转置成一列再重复增加到size(x)列。因此命令等效于：
X=ones(size(y))*x;
Y=y’*ones(size(x))
x = np.linspace(-1,1,2)
y = np.linspace(-1,1,5)
print(x)
print(y)
new_x, new_y = np.meshgrid(x,y)
print(new_x)
print(new_y)
'''
[-1. 1.]
[-1. -0.5 0. 0.5 1. ]
[[-1. 1.]
[-1. 1.]
[-1. 1.]
[-1. 1.]
[-1. 1.]]
[[-1. -1. ]
[-0.5 -0.5]
[ 0. 0. ]
[ 0.5 0.5]
[ 1. 1. ]]
'''numpy.vstack()
按列堆叠矩阵
a = np.identity(2)
b = np.identity(2)*2
c = np.vstack([a,b])
print(c)
'''
[[1. 0.]
[0. 1.]
[2. 0.]
[0. 2.]]
'''numpy.prod()
对应轴上元素的乘积 默认没有轴，即所有元素乘积。 axis=0 按列 axis=1 按行
a = np.prod([[1.,2.],[3.,4.]])
print(a)
a = np.prod([[1.,2.],[3.,4.]], axis=0)
print(a)
a = np.prod([[1.,2.],[3.,4.]], axis=1)
print(a)
'''
24.0
[3. 8.]
[ 2. 12.]
'''—————————————————————————————————————
tensorflow.range()
用于构造数列。
range(limit, delta=1, dtype=None, name='range')
range(start, limit, delta=1, dtype=None, name='range')tensorflow.tile()
通过平铺（tile）给定的张量来构造张量。第i维平铺multiples[i]次
tf.tile(
input,
multiples,
name=None
)with tf.Session() as sess:
a = tf.tile([1,2,3],[2])
print(sess.run(a))
b = tf.tile([[1,2,3],[4,5,6],[7,8,9]],[2,3])
print(sess.run(b))
'''
[1 2 3 1 2 3]
[[1 2 3 1 2 3 1 2 3]
[4 5 6 4 5 6 4 5 6]
[7 8 9 7 8 9 7 8 9]
[1 2 3 1 2 3 1 2 3]
[4 5 6 4 5 6 4 5 6]
[7 8 9 7 8 9 7 8 9]]
'''tensorflow.stack()
tf.stack(
values,
axis=0,
name='stack'
)x = tf.constant([1, 4])
y = tf.constant([2, 5])
z = tf.constant([3, 6])
tf.stack([x, y, z]) # [[1, 4], [2, 5], [3, 6]] (Pack along first dim.)
tf.stack([x, y, z], axis=1) # [[1, 2, 3], [4, 5, 6]]给定一个形状为(A, B, C)的张量的长度 N 的列表；
如果 axis == 0，那么 output 张量将具有形状(N, A, B, C)。如果 axis == 1，那么 output 张量将具有形状(A, N, B, C)。
a = tf.ones((2,3,4))
b = tf.zeros((2,3,4))
c = tf.ones((2,3,4)) * 2
indices = tf.stack([a, b, c])
'''
[[[[1. 1. 1. 1.]
[1. 1. 1. 1.]
[1. 1. 1. 1.]]
[[1. 1. 1. 1.]
[1. 1. 1. 1.]
[1. 1. 1. 1.]]]
[[[0. 0. 0. 0.]
[0. 0. 0. 0.]
[0. 0. 0. 0.]]
[[0. 0. 0. 0.]
[0. 0. 0. 0.]
[0. 0. 0. 0.]]]
[[[2. 2. 2. 2.]
[2. 2. 2. 2.]
[2. 2. 2. 2.]]
[[2. 2. 2. 2.]
[2. 2. 2. 2.]
[2. 2. 2. 2.]]]]
'''
tensorflow.gather_nd()
将参数中的切片收集到由索引指定的形状的张量中
gather_nd(
params,
indices,
name=None
)indices = [[0, 1], [1, 0]]
params = [[['a0', 'b0'], ['c0', 'd0']],
[['a1', 'b1'], ['c1', 'd1']]]
output = [['c0', 'd0'], ['a1', 'b1']]References:
https://blog.csdn.net/m0_37908327/article/details/68953784
https://www.cnblogs.com/sunshinewang/p/6897966.html