How Can A String Representation Of A NumPy Array Be Converted To A NumPy Array?

The function numpy.array_repr can be used to create a string representation of a NumPy array. How can a string representation of a NumPy array be converted to a NumPy array? Let's

Solution 1:

eval is the easiest, probably. It evaluates a given string as if it were code.

from numpy import array, all
arr_1 = array([1,2,3])
arr_string = repr(arr_1)
arr_2 = eval(arr_string)

all(arr_1 == arr_2) # True

See also documentation on eval: https://docs.python.org/2/library/functions.html#eval

Solution 2:

I often debug with print statements. To read numpy output from the console back into a python environment, I use the following utility based on np.matrix.

def string_to_numpy(text, dtype=None):
    """
    Convert text into 1D or 2D arrays using np.matrix().
    The result is returned as an np.ndarray.
    """
    import re
    text = text.strip()
    # Using a regexp, decide whether the array is flat or not.
    # The following matches either: "[1 2 3]" or "1 2 3"
    is_flat = bool(re.match(r"^(\[[^\[].+[^\]]\]|[^\[].+[^\]])$",
                            text, flags=re.S))
    # Replace newline characters with semicolons.
    text = text.replace("]\n", "];")
    # Prepare the result.
    result = np.asarray(np.matrix(text, dtype=dtype))
    return result.flatten() if is_flat else result

---

Here's the workflow that I often apply for debugging:

1) Somewhere in my code...

import numpy as np
x = np.random.random((3,5)).round(decimals=2)
print(x)

2. This prints the content of the array onto the console, for example:

    [[0.24 0.68 0.57 0.37 0.83]
     [0.76 0.5  0.46 0.49 0.95]
     [0.39 0.37 0.48 0.69 0.25]]

3. To further examine the output, I select the text and paste it in a ipython session as follows:

    In [9]: s2n = string_to_numpy # Short alias

    In [10]: x = s2n("""[[0.24 0.68 0.57 0.37 0.83]
                         [0.76 0.5  0.46 0.49 0.95]
                         [0.39 0.37 0.48 0.69 0.25]]""")
    In [11]: x.shape
    Out[11]: (3, 5)

    In [12]: x.mean(axis=1)
    Out[12]: array([0.538, 0.632, 0.436])
    
    ...