<p>如中所述</p>
<p><a href="https://stackoverflow.com/questions/35747614/loading-arrays-saved-using-numpy-save-in-append-mode">loading arrays saved using numpy.save in append mode</a></p>
<p>可以多次保存到打开的文件中,也可以多次加载。这并没有记录在案,也许也不是首选,但它确实有效。<code>savez</code>存档是保存多个数组的首选方法。</p>
<p>下面是一个玩具示例:</p>
<pre><code>In [777]: with open('multisave.npy','wb') as f:
...: arr = np.arange(10)
...: np.save(f, arr)
...: arr = np.arange(20)
...: np.save(f, arr)
...: arr = np.ones((3,4))
...: np.save(f, arr)
...:
In [778]: ll multisave.npy
-rw-rw-r-- 1 paul 456 Feb 13 08:38 multisave.npy
In [779]: with open('multisave.npy','rb') as f:
...: arr = np.load(f)
...: print(arr)
...: print(np.load(f))
...: print(np.load(f))
...:
[0 1 2 3 4 5 6 7 8 9]
[ 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19]
[[ 1. 1. 1. 1.]
[ 1. 1. 1. 1.]
[ 1. 1. 1. 1.]]
</code></pre>
<hr/>
<p>下面是一个保存相同形状数组列表的简单示例</p>
<pre><code>In [780]: traces = [np.arange(10),np.arange(10,20),np.arange(100,110)]
In [781]: traces
Out[781]:
[array([0, 1, 2, 3, 4, 5, 6, 7, 8, 9]),
array([10, 11, 12, 13, 14, 15, 16, 17, 18, 19]),
array([100, 101, 102, 103, 104, 105, 106, 107, 108, 109])]
In [782]: arr = np.array(traces)
In [783]: arr
Out[783]:
array([[ 0, 1, 2, 3, 4, 5, 6, 7, 8, 9],
[ 10, 11, 12, 13, 14, 15, 16, 17, 18, 19],
[100, 101, 102, 103, 104, 105, 106, 107, 108, 109]])
In [785]: np.save('mult1.npy', arr)
In [786]: data = np.load('mult1.npy')
In [787]: data
Out[787]:
array([[ 0, 1, 2, 3, 4, 5, 6, 7, 8, 9],
[ 10, 11, 12, 13, 14, 15, 16, 17, 18, 19],
[100, 101, 102, 103, 104, 105, 106, 107, 108, 109]])
In [788]: list(data)
Out[788]:
[array([0, 1, 2, 3, 4, 5, 6, 7, 8, 9]),
array([10, 11, 12, 13, 14, 15, 16, 17, 18, 19]),
array([100, 101, 102, 103, 104, 105, 106, 107, 108, 109])]
</code></pre>