Python 一个相当于熊猫阅读剪贴板的NumPy？

例如，如果您遇到的问题/答案发布了如下数组：

[[ 0  1  2  3  4  5  6  7]
 [ 8  9 10 11 12 13 14 15]
 [16 17 18 19 20 21 22 23]
 [24 25 26 27 28 29 30 31]
 [32 33 34 35 36 37 38 39]
 [40 41 42 43 44 45 46 47]
 [48 49 50 51 52 53 54 55]
 [56 57 58 59 60 61 62 63]]

from pandas.io.clipboard import clipboard_get
# import pyperclip
import numpy as np
import re
import ast

def numpy_from_clipboard():
    inp = clipboard_get()
    # inp = pyperclip.paste()
    inp = inp.strip()
    # if it starts with "array(" we just need to remove the
    # leading "array(" and remove the optional ", dtype=xxx)"
    if inp.startswith('array('):
        inp = re.sub(r'^array\(', '', inp)
        dtype = re.search(r', dtype=(\w+)\)$', inp)
        if dtype:
            return np.array(ast.literal_eval(inp[:dtype.start()]), dtype=dtype.group(1))
        else:
            return np.array(ast.literal_eval(inp[:-1]))
    else:
        # In case it's the string representation it's a bit harder.
        # We need to remove all spaces between closing and opening brackets
        inp = re.sub(r'\]\s+\[', '],[', inp)
        # We need to remove all whitespaces following an opening bracket
        inp = re.sub(r'\[\s+', '[', inp)
        # and all leading whitespaces before closing brackets
        inp = re.sub(r'\s+\]', ']', inp)
        # replace all remaining whitespaces with ","
        inp = re.sub(r'\s+', ',', inp)
        return np.array(ast.literal_eval(inp))

---

您如何在REPL会话中将其加载到变量中，而不必到处添加逗号？

对于一次性情况，我可能会这样做：

• 将包含数组的文本复制到剪贴板
• 在ipython shell中，输入

  s=“”

  ，但不要点击回车键
• 从剪贴板粘贴文本
• 键入结束的三引号

这给了我：

In [16]: s = """[[ 0  1  2  3  4  5  6  7]
    ...:  [ 8  9 10 11 12 13 14 15]
    ...:  [16 17 18 19 20 21 22 23]
    ...:  [24 25 26 27 28 29 30 31]
    ...:  [32 33 34 35 36 37 38 39]
    ...:  [40 41 42 43 44 45 46 47]
    ...:  [48 49 50 51 52 53 54 55]
    ...:  [56 57 58 59 60 61 62 63]]"""

然后使用

np.loadtxt（）

，如下所示：

In [17]: a = np.loadtxt([line.lstrip(' [').rstrip(']') for line in s.splitlines()], dtype=int)

In [18]: a
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],
       [24, 25, 26, 27, 28, 29, 30, 31],
       [32, 33, 34, 35, 36, 37, 38, 39],
       [40, 41, 42, 43, 44, 45, 46, 47],
       [48, 49, 50, 51, 52, 53, 54, 55],
       [56, 57, 58, 59, 60, 61, 62, 63]])

---

如果你有熊猫，pyperclip或者你可以用这样的东西：

[[ 0  1  2  3  4  5  6  7]
 [ 8  9 10 11 12 13 14 15]
 [16 17 18 19 20 21 22 23]
 [24 25 26 27 28 29 30 31]
 [32 33 34 35 36 37 38 39]
 [40 41 42 43 44 45 46 47]
 [48 49 50 51 52 53 54 55]
 [56 57 58 59 60 61 62 63]]

from pandas.io.clipboard import clipboard_get
# import pyperclip
import numpy as np
import re
import ast

def numpy_from_clipboard():
    inp = clipboard_get()
    # inp = pyperclip.paste()
    inp = inp.strip()
    # if it starts with "array(" we just need to remove the
    # leading "array(" and remove the optional ", dtype=xxx)"
    if inp.startswith('array('):
        inp = re.sub(r'^array\(', '', inp)
        dtype = re.search(r', dtype=(\w+)\)$', inp)
        if dtype:
            return np.array(ast.literal_eval(inp[:dtype.start()]), dtype=dtype.group(1))
        else:
            return np.array(ast.literal_eval(inp[:-1]))
    else:
        # In case it's the string representation it's a bit harder.
        # We need to remove all spaces between closing and opening brackets
        inp = re.sub(r'\]\s+\[', '],[', inp)
        # We need to remove all whitespaces following an opening bracket
        inp = re.sub(r'\[\s+', '[', inp)
        # and all leading whitespaces before closing brackets
        inp = re.sub(r'\s+\]', ']', inp)
        # replace all remaining whitespaces with ","
        inp = re.sub(r'\s+', ',', inp)
        return np.array(ast.literal_eval(inp))

然后阅读您保存在剪贴板中的内容：

>>> numpy_from_clipboard()
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],
       [24, 25, 26, 27, 28, 29, 30, 31],
       [32, 33, 34, 35, 36, 37, 38, 39],
       [40, 41, 42, 43, 44, 45, 46, 47],
       [48, 49, 50, 51, 52, 53, 54, 55],
       [56, 57, 58, 59, 60, 61, 62, 63]])

这应该能够从剪贴板解析（大多数）数组（

str

以及数组的

repr

）。它甚至可以用于多行数组（其中

np.loadtxt

失败）：

但是我不太擅长正则表达式，所以这可能不是万无一失的，使用

ast.literal\u eval

感觉有点笨拙（但它避免了自己进行解析）

---

请随意提出改进建议。

我认为没有直接对应的。@juanpa.arrivillaga好吧，应该有。超方便方法ftw.Ehhh我更认为人们应该负责提供可复制的示例。打印（repr（arr））有多困难？但这是一个很好的问题。@juanpa.arrivillaga 100%同意你的观点。遗憾的是，有可复制例子的问题是濒危物种，我尽可能地珍视它们。，