np.uniform用法、数据维度扩充用法示例

rng = np.random.RandomState(42)
X = rng.uniform(0, 5, 100)
X = X[:, np.newaxis]  #把行向量转化为了列向量
y = np.array(X[:, 0] > 2.5, dtype=int) #

等价于:

rng = np.random.RandomState(42)
X = rng.uniform(0, 5, [100,1])
X = X[:, np.newaxis]
y = np.array(X[:, 0] > 2.5, dtype=int)

验证一下：

>>> import numpy as np
>>> rng = np.random.RandomState(42)
>>> X = rng.uniform(0,5,5)
>>> print(X)
[1.87270059 4.75357153 3.65996971 2.99329242 0.7800932 ]
>>> print(X[:,np.newaxis])
[[1.87270059]
 [4.75357153]
 [3.65996971]
 [2.99329242]
 [0.7800932 ]]

如果需要扩充数据的维度，可以用np.newaxis，验证：

>>> X = rng.uniform(1,5,3)
>>> print(X[:,np.newaxis,np.newaxis])
[[[1.08233798]]

 [[4.87963941]]

 [[4.32977056]]]