python - 如何添加到 numpy 数组的末尾并从头开始删除？

Question

我有一个形状为 (1, 100) 的数组，即:

[[1. 2. 3. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.
  0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.
  0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.
  0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.
  0. 0. 0. 55.]]

我想在末尾添加一些东西(例如 123 之类的数字)并删除第一个元素，这样我将拥有:

[[2. 3. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.
  0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.
  0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.
  0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.
  0. 0. 55. 123.]]

其中保留了原来的形状(1, 100)

我正在尝试:

x_pred = np.append(x_pred, next_index, axis=1)

(x_pred 是一个(1, 100) 数组，next_index 是一个标量)

但是我得到一个错误:

ValueError: all the input arrays must have same number of dimensions

我做错了什么？

Answer 1

您可以使用 roll 执行此操作。

a = np.zeros((1,10))

#roll and replace
a[0] = np.roll(a[0],-1)
a[0][-1] = new_value