entityMap|blocks|key|7n7qk|text|根据|type|unstyled|depth|inlineStyleRanges|entityRanges|data|mqum|newshape+:+int或int的元组|js1c|新形状应与原始形状兼容。如果是整数，则结果将是该长度的一维数组。一个形状维度可以是|4sus4|-1。在这种情况下，该值是从数组的长度和剩余维度推断出来的。^0|0|0|0^^$0|$]|1|@$2|3|4|5|6|7|8|I|9|@]|A|@]|B|$]]|$2|C|4|D|6|7|8|J|9|@]|A|@]|B|$]]|$2|E|4|F|6|7|8|K|9|@]|A|@]|B|$]]|$2|G|4|H|6|7|8|L|9|@]|A|@]|B|$]]]]

According to <a href="http://docs.scipy.org/doc/numpy/reference/generated/numpy.reshape.html#numpy-reshape"><code>the documentation</code></a>:

<blockquote>
 newshape : int or tuple of ints
 
 The new shape should be compatible with the original shape. If an
 integer, then the result will be a 1-D array of that length. One shape
 dimension can be -1. In this case, the value is inferred from the
 length of the array and remaining dimensions.
</blockquote>

entityMap|blocks|key|3cmbl|text|numpy.reshape(a,newshape,order{})|type|code-block|depth|inlineStyleRanges|entityRanges|data|syntax|javascript|5r0qf|查看下面的链接了解更多信息。|unstyled|2sjk1|https://docs.scipy.org/doc/numpy/reference/generated/numpy.reshape.html|brm3t|对于您提到的下面的示例，输出将结果向量解释为单行。(-+1+)表示行数为1。|8s7nb|a+=+numpy.matrix([[1,+2,+3,+4],+[5,+6,+7,+8]])
b+=+numpy.reshape(a,+-1)|7neug|输出：|7g040|matrix([[1,+2,+3,+4,+5,+6,+7,+8]])|3gksv|可以用另一个例子更准确地解释这一点：|cpo9n|b+=+np.arange(10).reshape((-1,1))|9qcis|输出：(一维数组)|e826q|array([[0],
+++++++[1],
+++++++[2],
+++++++[3],
+++++++[4],
+++++++[5],
+++++++[6],
+++++++[7],
+++++++[8],
+++++++[9]])|8dg1k|或者|5ndi6|b+=+np.arange(10).reshape((1,-1))|9hioq|输出：(是一个一维的行数组)|9tpcg|array([[0,+1,+2,+3,+4,+5,+6,+7,+8,+9]])^0|0|0|0|0|0|0|0|0|0|0|0|0|0|0^^$0|$]|1|@$2|3|4|5|6|7|8|17|9|@]|A|@]|B|$C|D]]|$2|E|4|F|6|G|8|18|9|@]|A|@]|B|$]]|$2|H|4|I|6|G|8|19|9|@]|A|@]|B|$]]|$2|J|4|K|6|G|8|1A|9|@]|A|@]|B|$]]|$2|L|4|M|6|7|8|1B|9|@]|A|@]|B|$C|D]]|$2|N|4|O|6|G|8|1C|9|@]|A|@]|B|$]]|$2|P|4|Q|6|7|8|1D|9|@]|A|@]|B|$C|D]]|$2|R|4|S|6|G|8|1E|9|@]|A|@]|B|$]]|$2|T|4|U|6|7|8|1F|9|@]|A|@]|B|$C|D]]|$2|V|4|W|6|G|8|1G|9|@]|A|@]|B|$]]|$2|X|4|Y|6|7|8|1H|9|@]|A|@]|B|$C|D]]|$2|Z|4|10|6|G|8|1I|9|@]|A|@]|B|$]]|$2|11|4|12|6|7|8|1J|9|@]|A|@]|B|$C|D]]|$2|13|4|14|6|G|8|1K|9|@]|A|@]|B|$]]|$2|15|4|16|6|7|8|1L|9|@]|A|@]|B|$C|D]]]]

<pre><code>numpy.reshape(a,newshape,order{})
</code></pre>
check the below link for more info.
<a href="https://docs.scipy.org/doc/numpy/reference/generated/numpy.reshape.html" rel="nofollow noreferrer">https://docs.scipy.org/doc/numpy/reference/generated/numpy.reshape.html</a>
for the below example you mentioned the output explains the resultant vector to be a single row.(-1) indicates the number of rows to be 1.
if the
<pre><code>a = numpy.matrix([[1, 2, 3, 4], [5, 6, 7, 8]])
b = numpy.reshape(a, -1)
</code></pre>
output:
<pre><code>matrix([[1, 2, 3, 4, 5, 6, 7, 8]])
</code></pre>
this can be explained more precisely with another example:
<pre><code>b = np.arange(10).reshape((-1,1))
</code></pre>
output:(is a 1 dimensional columnar array)
<pre><code>array([[0],
 [1],
 [2],
 [3],
 [4],
 [5],
 [6],
 [7],
 [8],
 [9]])
</code></pre>
or
<pre><code>b = np.arange(10).reshape((1,-1))
</code></pre>
output:(is a 1 dimensional row array)
<pre><code>array([[0, 1, 2, 3, 4, 5, 6, 7, 8, 9]])
</code></pre>

entityMap|blocks|key|18h89|text|这是相当容易理解的。"-1“代表”未知维度“，可以从另一个维度推断。在这种情况下，如果您像这样设置矩阵：|type|unstyled|depth|inlineStyleRanges|entityRanges|data|c10ne|a+=+numpy.matrix([[1,+2,+3,+4],+[5,+6,+7,+8]])|code-block|syntax|javascript|a6r8o|如下所示修改矩阵：|coiqf|b+=+numpy.reshape(a,+-1)|ctgg1|它将调用一些对矩阵a的默认操作，这将返回一个一维numpy数组/矩阵。|7r1f|但是，我不认为使用这样的代码是一个好主意。为什么不试试：|3or3f|b+=+a.reshape(1,-1)|97k02|对于a，我们不知道它应该有多少列(将其设置为-+1!)，但我们需要一个一维数组(将第一个参数设置为1！)。^0|0|0|0|0|0|0|0^^$0|$]|1|@$2|3|4|5|6|7|8|T|9|@]|A|@]|B|$]]|$2|C|4|D|6|E|8|U|9|@]|A|@]|B|$F|G]]|$2|H|4|I|6|7|8|V|9|@]|A|@]|B|$]]|$2|J|4|K|6|E|8|W|9|@]|A|@]|B|$F|G]]|$2|L|4|M|6|7|8|X|9|@]|A|@]|B|$]]|$2|N|4|O|6|7|8|Y|9|@]|A|@]|B|$]]|$2|P|4|Q|6|E|8|Z|9|@]|A|@]|B|$F|G]]|$2|R|4|S|6|7|8|10|9|@]|A|@]|B|$]]]]

The <code>-1</code> stands for &quot;unknown dimension&quot; which can be inferred from another dimension.
In this case, if you set your matrix like this:
<pre><code>a = numpy.matrix([[1, 2, 3, 4], [5, 6, 7, 8]])
</code></pre>
Modify your matrix like this:
<pre><code>b = numpy.reshape(a, -1)
</code></pre>
It will call some default operations to the matrix <code>a</code>, which will return a 1-d <code>numpy</code> array/matrix.
However, I don't think it is a good idea to use code like this. Why not try:
<pre><code>b = a.reshape(1, -1)
</code></pre>
It will give you the same result and it's more clear for readers to understand: Set <code>b</code> as another shape of <code>a</code>. For <code>a</code>, we don't how many columns it should have (set it to -1!), but we want a 1-dimension array (set the first parameter to 1!).

entityMap|blocks|key|b16f3|text|提供新形状要满足的标准是|type|unstyled|depth|inlineStyleRanges|entityRanges|data|77gqt|‘新形状应与原始形状兼容’|7g3bq|numpy允许我们给出一个新的形状参数为-1+(例如：(2，-1)或(-1,3)，而不是(-1，-1))。这仅仅意味着它是一个未知的维度，我们想让numpy来解决它。numpy将通过查看|brii6|‘数组的长度和剩余维度’|6qtgq|并确保它满足上述标准|4olit|现在请看示例。|5lluc|z+=+np.array([[1,+2,+3,+4],
+++++++++[5,+6,+7,+8],
+++++++++[9,+10,+11,+12]])
z.shape
(3,+4)|code-block|syntax|javascript|duegl|现在尝试使用(-1)重塑。结果新形状为(12，)，并与原始形状(3,4)兼容|4eugt|z.reshape(-1)
array([+1,++2,++3,++4,++5,++6,++7,++8,++9,+10,+11,+12])|ai6p2|现在尝试用(-1，1)重塑。我们提供了列作为1，但行作为未知。因此我们得到的新形状为(12，1).again与原始形状(3，4)兼容|8gh4a|z.reshape(-1,1)
array([[+1],
+++[+2],
+++[+3],
+++[+4],
+++[+5],
+++[+6],
+++[+7],
+++[+8],
+++[+9],
+++[10],
+++[11],
+++[12]])|bgnla|以上内容与|c3vqb|建议/错误消息，使用|9iv4u|用于单个功能；即单列|5236f|使用重塑您的数据|ejt22|如果您的数据具有|5d5l5|单一功能|8nkrt|1725g|3544a|新形状为(-1，2)。行未知，列2。我们得到的结果新形状为(6，2)。|6c1vk|z.reshape(-1,+2)
array([[+1,++2],
+++[+3,++4],
+++[+5,++6],
+++[+7,++8],
+++[+9,+10],
+++[11,+12]])|12hem|现在尝试将列保留为未知。新形状为(1，-1)。即，行为1，列未知。我们得到的结果新形状为(1，12)|9tq2t|z.reshape(1,-1)
array([[+1,++2,++3,++4,++5,++6,++7,++8,++9,+10,+11,+12]])|9kpll|vlmr|crutj|对于单个样本；即单行|feire|fel44|如果它包含一个|3c8n7|单一样本|1upqk|djfh9|34ppg|新形状(2，-1)。第2行，第2列未知。我们得到的结果是新的形状为(2,6)|2pis0|z.reshape(2,+-1)
array([[+1,++2,++3,++4,++5,++6],
+++[+7,++8,++9,+10,+11,+12]])|kgpm|新形状为(3，-1)。第3行，列未知。我们得到的结果是新的形状为(3,4)|9hlig|z.reshape(3,+-1)
array([[+1,++2,++3,++4],
+++[+5,++6,++7,++8],
+++[+9,+10,+11,+12]])|6ndj9|最后，如果我们尝试将两个维度都提供为未知，即新形状为(-1，-1)。它将抛出一个错误|5jr4p|z.reshape(-1,+-1)
ValueError:+can+only+specify+one+unknown+dimension^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^^$0|$]|1|@$2|3|4|5|6|7|8|28|9|@]|A|@]|B|$]]|$2|C|4|D|6|7|8|29|9|@]|A|@]|B|$]]|$2|E|4|F|6|7|8|2A|9|@]|A|@]|B|$]]|$2|G|4|H|6|7|8|2B|9|@]|A|@]|B|$]]|$2|I|4|J|6|7|8|2C|9|@]|A|@]|B|$]]|$2|K|4|L|6|7|8|2D|9|@]|A|@]|B|$]]|$2|M|4|N|6|O|8|2E|9|@]|A|@]|B|$P|Q]]|$2|R|4|S|6|7|8|2F|9|@]|A|@]|B|$]]|$2|T|4|U|6|O|8|2G|9|@]|A|@]|B|$P|Q]]|$2|V|4|W|6|7|8|2H|9|@]|A|@]|B|$]]|$2|X|4|Y|6|O|8|2I|9|@]|A|@]|B|$P|Q]]|$2|Z|4|10|6|7|8|2J|9|@]|A|@]|B|$]]|$2|11|4|12|6|7|8|2K|9|@]|A|@]|B|$]]|$2|13|4|14|6|7|8|2L|9|@]|A|@]|B|$]]|$2|15|4|16|6|7|8|2M|9|@]|A|@]|B|$]]|$2|17|4|18|6|7|8|2N|9|@]|A|@]|B|$]]|$2|19|4|1A|6|7|8|2O|9|@]|A|@]|B|$]]|$2|1B|4|-4|6|7|8|2P|9|@]|A|@]|B|$]]|$2|1C|4|-4|6|7|8|2Q|9|@]|A|@]|B|$]]|$2|1D|4|1E|6|7|8|2R|9|@]|A|@]|B|$]]|$2|1F|4|1G|6|O|8|2S|9|@]|A|@]|B|$P|Q]]|$2|1H|4|1I|6|7|8|2T|9|@]|A|@]|B|$]]|$2|1J|4|1K|6|O|8|2U|9|@]|A|@]|B|$P|Q]]|$2|1L|4|10|6|7|8|2V|9|@]|A|@]|B|$]]|$2|1M|4|12|6|7|8|2W|9|@]|A|@]|B|$]]|$2|1N|4|1O|6|7|8|2X|9|@]|A|@]|B|$]]|$2|1P|4|16|6|7|8|2Y|9|@]|A|@]|B|$]]|$2|1Q|4|1R|6|7|8|2Z|9|@]|A|@]|B|$]]|$2|1S|4|1T|6|7|8|30|9|@]|A|@]|B|$]]|$2|1U|4|-4|6|7|8|31|9|@]|A|@]|B|$]]|$2|1V|4|-4|6|7|8|32|9|@]|A|@]|B|$]]|$2|1W|4|1X|6|7|8|33|9|@]|A|@]|B|$]]|$2|1Y|4|1Z|6|O|8|34|9|@]|A|@]|B|$P|Q]]|$2|20|4|21|6|7|8|35|9|@]|A|@]|B|$]]|$2|22|4|23|6|O|8|36|9|@]|A|@]|B|$P|Q]]|$2|24|4|25|6|7|8|37|9|@]|A|@]|B|$]]|$2|26|4|27|6|O|8|38|9|@]|A|@]|B|$P|Q]]]]

The criterion to satisfy for providing the new shape is that 'The new shape should be compatible with the original shape'

numpy allow us to give one of new shape parameter as -1 (eg: (2,-1) or (-1,3) but not (-1, -1)). It simply means that it is an unknown dimension and we want numpy to figure it out. And numpy will figure this by looking at the 'length of the array and remaining dimensions' and making sure it satisfies the above mentioned criteria

Now see the example.

<pre><code>z = np.array([[1, 2, 3, 4],
 [5, 6, 7, 8],
 [9, 10, 11, 12]])
z.shape
(3, 4)
</code></pre>

Now trying to reshape with (-1) . Result new shape is (12,) and is compatible with original shape (3,4) 

<pre><code>z.reshape(-1)
array([ 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12])
</code></pre>

Now trying to reshape with (-1, 1) . We have provided column as 1 but rows as unknown . So we get result new shape as (12, 1).again compatible with original shape(3,4) 

<pre><code>z.reshape(-1,1)
array([[ 1],
 [ 2],
 [ 3],
 [ 4],
 [ 5],
 [ 6],
 [ 7],
 [ 8],
 [ 9],
 [10],
 [11],
 [12]])
</code></pre>

The above is consistent with <code>numpy</code> advice/error message, to use <code>reshape(-1,1)</code> for a single feature; i.e. single column

<blockquote>
 Reshape your data using <code>array.reshape(-1, 1)</code> if your data has a single feature
</blockquote>

New shape as (-1, 2). row unknown, column 2. we get result new shape as (6, 2)

<pre><code>z.reshape(-1, 2)
array([[ 1, 2],
 [ 3, 4],
 [ 5, 6],
 [ 7, 8],
 [ 9, 10],
 [11, 12]])
</code></pre>

Now trying to keep column as unknown. New shape as (1,-1). i.e, row is 1, column unknown. we get result new shape as (1, 12)

<pre><code>z.reshape(1,-1)
array([[ 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12]])
</code></pre>

The above is consistent with <code>numpy</code> advice/error message, to use <code>reshape(1,-1)</code> for a single sample; i.e. single row

<blockquote>
 Reshape your data using <code>array.reshape(1, -1)</code> if it contains a single sample
</blockquote>

New shape (2, -1). Row 2, column unknown. we get result new shape as (2,6)

<pre><code>z.reshape(2, -1)
array([[ 1, 2, 3, 4, 5, 6],
 [ 7, 8, 9, 10, 11, 12]])
</code></pre>

New shape as (3, -1). Row 3, column unknown. we get result new shape as (3,4)

<pre><code>z.reshape(3, -1)
array([[ 1, 2, 3, 4],
 [ 5, 6, 7, 8],
 [ 9, 10, 11, 12]])
</code></pre>

And finally, if we try to provide both dimension as unknown i.e new shape as (-1,-1). It will throw an error

<pre><code>z.reshape(-1, -1)
ValueError: can only specify one unknown dimension
</code></pre>

entityMap|blocks|key|3fghm|text|长话短说|type|unstyled|depth|inlineStyleRanges|entityRanges|data|ejk5q|：您设置了一些尺寸，并让NumPy设置其余的尺寸。|54urk|(userDim1,+userDim2,+...,+-1)+-->>

(userDim1,+userDim1,+...,+TOTAL_DIMENSION+-+(userDim1+%2B+userDim2+%2B+...))|code-block|syntax|javascript^0|0|0^^$0|$]|1|@$2|3|4|5|6|7|8|J|9|@]|A|@]|B|$]]|$2|C|4|D|6|7|8|K|9|@]|A|@]|B|$]]|$2|E|4|F|6|G|8|L|9|@]|A|@]|B|$H|I]]]]

Long story short: you set some dimensions and let NumPy set the remaining(s).

<pre><code>(userDim1, userDim2, ..., -1) --&gt;&gt;

(userDim1, userDim1, ..., TOTAL_DIMENSION - (userDim1 + userDim2 + ...))
</code></pre>

entityMap|blocks|key|49hnp|text|这仅仅意味着您不确定您可以给出多少行或列数，并且您要求numpy建议要重塑的列数或行数。|type|unstyled|depth|inlineStyleRanges|entityRanges|data|4jte7|numpy提供了-1的最后一个示例|pllp|https://docs.scipy.org/doc/numpy/reference/generated/numpy.reshape.html|684ep|请查看以下代码及其输出，以更好地了解(-1)：|a8jns|代码：-|2egf0|import+numpy
a+=+numpy.matrix([[1,+2,+3,+4],+[5,+6,+7,+8]])
print("Without+reshaping++->+")
print(a)
b+=+numpy.reshape(a,+-1)
print("HERE+We+don't+know+about+what+number+we+should+give+to+row/col")
print("Reshaping+as+(a,-1)")
print(b)
c+=+numpy.reshape(a,+(-1,2))
print("HERE+We+just+know+about+number+of+columns")
print("Reshaping+as+(a,(-1,2))")
print(c)
d+=+numpy.reshape(a,+(2,-1))
print("HERE+We+just+know+about+number+of+rows")
print("Reshaping+as+(a,(2,-1))")
print(d)|code-block|syntax|javascript|9u8af|输出:-|1g40u|Without+reshaping++->+
[[1+2+3+4]
+[5+6+7+8]]
HERE+We+don't+know+about+what+number+we+should+give+to+row/col
Reshaping+as+(a,-1)
[[1+2+3+4+5+6+7+8]]
HERE+We+just+know+about+number+of+columns
Reshaping+as+(a,(-1,2))
[[1+2]
+[3+4]
+[5+6]
+[7+8]]
HERE+We+just+know+about+number+of+rows
Reshaping+as+(a,(2,-1))
[[1+2+3+4]
+[5+6+7+8]]^0|0|0|0|0|0|0|0^^$0|$]|1|@$2|3|4|5|6|7|8|T|9|@]|A|@]|B|$]]|$2|C|4|D|6|7|8|U|9|@]|A|@]|B|$]]|$2|E|4|F|6|7|8|V|9|@]|A|@]|B|$]]|$2|G|4|H|6|7|8|W|9|@]|A|@]|B|$]]|$2|I|4|J|6|7|8|X|9|@]|A|@]|B|$]]|$2|K|4|L|6|M|8|Y|9|@]|A|@]|B|$N|O]]|$2|P|4|Q|6|7|8|Z|9|@]|A|@]|B|$]]|$2|R|4|S|6|M|8|10|9|@]|A|@]|B|$N|O]]]]

It simply means that you are not sure about what number of rows or columns you can give and you are asking numpy to suggest a number of column or rows to get reshaped in.
numpy provides the last example for -1
<a href="https://docs.scipy.org/doc/numpy/reference/generated/numpy.reshape.html" rel="nofollow noreferrer">https://docs.scipy.org/doc/numpy/reference/generated/numpy.reshape.html</a>
check below code and its output to better understand about (-1):
CODE:-
<pre><code>import numpy
a = numpy.matrix([[1, 2, 3, 4], [5, 6, 7, 8]])
print(&quot;Without reshaping -&gt; &quot;)
print(a)
b = numpy.reshape(a, -1)
print(&quot;HERE We don't know about what number we should give to row/col&quot;)
print(&quot;Reshaping as (a,-1)&quot;)
print(b)
c = numpy.reshape(a, (-1,2))
print(&quot;HERE We just know about number of columns&quot;)
print(&quot;Reshaping as (a,(-1,2))&quot;)
print(c)
d = numpy.reshape(a, (2,-1))
print(&quot;HERE We just know about number of rows&quot;)
print(&quot;Reshaping as (a,(2,-1))&quot;)
print(d)
</code></pre>
OUTPUT:-
<pre><code>Without reshaping -&gt; 
[[1 2 3 4]
 [5 6 7 8]]
HERE We don`t know about what number we should give to row/col
Reshaping as (a,-1)
[[1 2 3 4 5 6 7 8]]
HERE We just know about number of columns
Reshaping as (a,(-1,2))
[[1 2]
 [3 4]
 [5 6]
 [7 8]]
HERE We just know about number of rows
Reshaping as (a,(2,-1))
[[1 2 3 4]
 [5 6 7 8]]
</code></pre>

entityMap|blocks|key|dnte|text|import+numpy+as+np
x+=+np.array([[2,3,4],+[5,6,7]])+

#+Convert+any+shape+to+1D+shape
x+=+np.reshape(x,+(-1))+#+Making+it+1+row+->+(6,)

#+When+you+don't+care+about+rows+and+just+want+to+fix+number+of+columns
x+=+np.reshape(x,+(-1,+1))+#+Making+it+1+column+->+(6,+1)
x+=+np.reshape(x,+(-1,+2))+#+Making+it+2+column+->+(3,+2)
x+=+np.reshape(x,+(-1,+3))+#+Making+it+3+column+->+(2,+3)

#+When+you+don't+care+about+columns+and+just+want+to+fix+number+of+rows
x+=+np.reshape(x,+(1,+-1))+#+Making+it+1+row+->+(1,+6)
x+=+np.reshape(x,+(2,+-1))+#+Making+it+2+row+->+(2,+3)
x+=+np.reshape(x,+(3,+-1))+#+Making+it+3+row+->+(3,+2)|type|code-block|depth|inlineStyleRanges|entityRanges|data|syntax|javascript^0^^$0|$]|1|@$2|3|4|5|6|7|8|E|9|@]|A|@]|B|$C|D]]]]

<pre><code>import numpy as np
x = np.array([[2,3,4], [5,6,7]]) 

# Convert any shape to 1D shape
x = np.reshape(x, (-1)) # Making it 1 row -&gt; (6,)

# When you don't care about rows and just want to fix number of columns
x = np.reshape(x, (-1, 1)) # Making it 1 column -&gt; (6, 1)
x = np.reshape(x, (-1, 2)) # Making it 2 column -&gt; (3, 2)
x = np.reshape(x, (-1, 3)) # Making it 3 column -&gt; (2, 3)

# When you don't care about columns and just want to fix number of rows
x = np.reshape(x, (1, -1)) # Making it 1 row -&gt; (1, 6)
x = np.reshape(x, (2, -1)) # Making it 2 row -&gt; (2, 3)
x = np.reshape(x, (3, -1)) # Making it 3 row -&gt; (3, 2)
</code></pre>

entityMap|blocks|key|98rn3|text|转换的最终结果是最终数组中的元素数与初始数组或数据帧中的元素数相同。|type|unstyled|depth|inlineStyleRanges|entityRanges|data|7lr2m|对应于行或列的未知计数。我们可以把它想象成|4rktn|(未知)。|el5t3|是通过将原始数组中的元素数除以有序对的另一个值用-1得到的。|8goh8|示例：|ae664|12个元素|7890o|对应于具有|e8n17|=12/1=12行1列。|ak5sn|​|djf94|3blc4|对应于具有1行的数组，并且|5j83h|=12/1=12列。^0|0|0|0|0|0|0|0|0|0|0|0^^$0|$]|1|@$2|3|4|5|6|7|8|X|9|@]|A|@]|B|$]]|$2|C|4|D|6|7|8|Y|9|@]|A|@]|B|$]]|$2|E|4|F|6|7|8|Z|9|@]|A|@]|B|$]]|$2|G|4|H|6|7|8|10|9|@]|A|@]|B|$]]|$2|I|4|J|6|7|8|11|9|@]|A|@]|B|$]]|$2|K|4|L|6|7|8|12|9|@]|A|@]|B|$]]|$2|M|4|N|6|7|8|13|9|@]|A|@]|B|$]]|$2|O|4|P|6|7|8|14|9|@]|A|@]|B|$]]|$2|Q|4|R|6|7|8|15|9|@]|A|@]|B|$]]|$2|S|4|L|6|7|8|16|9|@]|A|@]|B|$]]|$2|T|4|U|6|7|8|17|9|@]|A|@]|B|$]]|$2|V|4|W|6|7|8|18|9|@]|A|@]|B|$]]]]

The final outcome of the conversion is that the number of elements in the final array is same as that of the initial array or data frame.
-1 corresponds to the unknown count of the row or column.
We can think of it as <code>x</code>(unknown). <code>x</code> is obtained by dividing the number of elements in the original array by the other value of the ordered pair with -1.
Examples:
12 elements with <code>reshape(-1,1)</code> corresponds to an array with <code>x</code>=12/1=12 rows and 1 column.
 
12 elements with <code>reshape(1,-1)</code> corresponds to an array with 1 row and <code>x</code>=12/1=12 columns.

entityMap|blocks|key|fj6pc|text|我不明白是什么|type|unstyled|depth|inlineStyleRanges|entityRanges|data|fei91|直到我读到|em1uh|本文|cag5c|..。|3q0f9|从机械上讲，这是很清楚的|eobqv|确实如此。但是我们如何解释重塑之前和之后的数据呢？|cq3jl|对我来说，缺少的部分是：|3mu4g|当我们训练机器学习模型时，数组的嵌套级别具有精确的含义。|dd56j|b2tf|这意味着整形操作|1bno5|必须敏锐地意识到|6sa0q|以下两点在操作前有任何意义：|48p4r|它所操作的数据(重塑输入看起来是什么样子)|6h6m8|算法/模型期望重塑的数据是什么样子(重塑输出是什么样子)|6dfa0|例如：|act37|外部数组包含观察值/行。内部数组包含列/特征。当我们只有一个特征的多个观察值的数组或多个特征的单个观察值时，这会导致两种特殊情况。|ag6ev|a7m49|有关更高级的示例:请参见|230fq|此堆栈溢出问题^0|0|0|0|0|0|0|0|0|0|0|0|0|0|0|0|0|0|0^^$0|$]|1|@$2|3|4|5|6|7|8|1A|9|@]|A|@]|B|$]]|$2|C|4|D|6|7|8|1B|9|@]|A|@]|B|$]]|$2|E|4|F|6|7|8|1C|9|@]|A|@]|B|$]]|$2|G|4|H|6|7|8|1D|9|@]|A|@]|B|$]]|$2|I|4|J|6|7|8|1E|9|@]|A|@]|B|$]]|$2|K|4|L|6|7|8|1F|9|@]|A|@]|B|$]]|$2|M|4|N|6|7|8|1G|9|@]|A|@]|B|$]]|$2|O|4|P|6|7|8|1H|9|@]|A|@]|B|$]]|$2|Q|4|-4|6|7|8|1I|9|@]|A|@]|B|$]]|$2|R|4|S|6|7|8|1J|9|@]|A|@]|B|$]]|$2|T|4|U|6|7|8|1K|9|@]|A|@]|B|$]]|$2|V|4|W|6|7|8|1L|9|@]|A|@]|B|$]]|$2|X|4|Y|6|7|8|1M|9|@]|A|@]|B|$]]|$2|Z|4|10|6|7|8|1N|9|@]|A|@]|B|$]]|$2|11|4|12|6|7|8|1O|9|@]|A|@]|B|$]]|$2|13|4|14|6|7|8|1P|9|@]|A|@]|B|$]]|$2|15|4|-4|6|7|8|1Q|9|@]|A|@]|B|$]]|$2|16|4|17|6|7|8|1R|9|@]|A|@]|B|$]]|$2|18|4|19|6|7|8|1S|9|@]|A|@]|B|$]]]]

I didn't manage to understand what <code>np.reshape()</code> does until I read <a href="https://www.mikulskibartosz.name/numpy-reshape-explained/" rel="nofollow noreferrer">this article</a>.
Mechanically it is clear what <code>reshape()</code> does. But how do we interpret the data before and after reshape?
The missing piece for me was:
<blockquote>
When we train a machine learning model, the nesting levels of arrays have precisely defined meaning.
</blockquote>
This means that the reshape operation has to be keenly aware both points below before the operation has any meaning:
<ul>
<li>the data it operates on (how the reshape input looks like)</li>
<li>how the algorithm/model expects the reshaped data to be (how the reshape output looks like)</li>
</ul>
For example:
<blockquote>
The external array contains observations/rows. The inner array contains columns/features. This causes two special cases when we have either an array of multiple observations of only one feature or a single observation of multiple features.
</blockquote>
For more advanced example:
See <a href="https://stackoverflow.com/questions/41396970/tensorflow-3-channel-order-of-color-inputs">this stackoverflow question</a>

A numpy matrix can be reshaped into a vector using reshape function with parameter -1. But I don't know what -1 means here.

For example: 

<pre><code>a = numpy.matrix([[1, 2, 3, 4], [5, 6, 7, 8]])
b = numpy.reshape(a, -1)
</code></pre>

The result of <code>b</code> is: <code>matrix([[1, 2, 3, 4, 5, 6, 7, 8]])</code>

Does anyone know what -1 means here?
And it seems python assign -1 several meanings, such as: <code>array[-1]</code> means the last element. Can you give an explanation?

What does -1 mean in numpy reshape?

机器学习

 可以使用参数为-1的重塑函数将数值矩阵重塑为向量。但我不知道-1在这里是什么意思。 例如： a = numpy.matrix([[1, 2, 3, 4], [5, 6, 7, 8]])b = numpy.reshape(a, -1) 的结果 是： 有人知道-1在这里是什么意思吗？python assign -1似乎有...

问-1在numpy重塑中是什么意思？
EN

社区

活动

资源

关于

腾讯云开发者

热门产品

热门推荐

更多推荐

问-1在numpy重塑中是什么意思？EN

社区

活动

资源

关于

腾讯云开发者

热门产品

热门推荐

更多推荐

问-1在numpy重塑中是什么意思？
EN