How to Convert a Python Sequence to a NumPy Array with Missing Values Filled?

Convert Python Sequence to NumPy Array with Missing Values Filled

When converting a Python sequence of variable-length lists to a NumPy array, an implicit conversion to an object type array occurs.

<code class="python">v = [[1], [1, 2]]
np.array(v)</code>

Output:

array([[1], [1, 2]], dtype=object)

Enforcing a specific data type, such as int32, will result in an exception:

<code class="python">np.array(v, dtype=np.int32)</code>

Exception:

ValueError: setting an array element with a sequence.

To obtain a dense NumPy array of int32 type with missing values filled with a placeholder, you can utilize itertools.zip_longest:

<code class="python">import itertools
np.array(list(itertools.zip_longest(*v, fillvalue=0))).T</code>

Output:

array([[1, 0],
       [1, 2]])

Note that in Python 2, itertools.izip_longest should be used instead.

The above is the detailed content of How to Convert a Python Sequence to a NumPy Array with Missing Values Filled?. For more information, please follow other related articles on the PHP Chinese website!