Python numpy 之argsort函数_numpy array argsort

argsort用于返回对数组排序之后，下表的交换情况

import numpy as np
x=np.array([1,4,3,-1,6,9])
y=x.argsort()
>>> y
array([3, 0, 2, 1, 4, 5], dtype=int64)

argsort(a, axis=-1, kind='quicksort', order=None)
    Returns the indices that would sort an array.
    
    Perform an indirect sort along the given axis using the algorithm specified
    by the `kind` keyword. It returns an array of indices of the same shape as
    `a` that index data along the given axis in sorted order.
    
    Parameters
    ----------
    a : array_like
        Array to sort.
    axis : int or None, optional
        Axis along which to sort.  The default is -1 (the last axis). If None,
        the flattened array is used.
    kind : {'quicksort', 'mergesort', 'heapsort'}, optional
        Sorting algorithm.
    order : str or list of str, optional
        When `a` is an array with fields defined, this argument specifies
        which fields to compare first, second, etc.  A single field can
        be specified as a string, and not all fields need be specified,
        but unspecified fields will still be used, in the order in which
        they come up in the dtype, to break ties.
    
    Returns
    -------
    index_array : ndarray, int
        Array of indices that sort `a` along the specified axis.
        If `a` is one-dimensional, ``a[index_array]`` yields a sorted `a`.
    
    See Also
    --------
    sort : Describes sorting algorithms used.
    lexsort : Indirect stable sort with multiple keys.
    ndarray.sort : Inplace sort.
    argpartition : Indirect partial sort.
    
    Notes
    -----
    See `sort` for notes on the different sorting algorithms.
    
    As of NumPy 1.4.0 `argsort` works with real/complex arrays containing
    nan values. The enhanced sort order is documented in `sort`.
    
    Examples
    --------
    One dimensional array:
    
    >>> x = np.array([3, 1, 2])
    >>> np.argsort(x)
    array([1, 2, 0])
    
    Two-dimensional array:
    
    >>> x = np.array([[0, 3], [2, 2]])
    >>> x
    array([[0, 3],
           [2, 2]])
    
    >>> np.argsort(x, axis=0)
    array([[0, 1],
           [1, 0]])
    
    >>> np.argsort(x, axis=1)
    array([[0, 1],
           [0, 1]])
    
    Sorting with keys:
    
    >>> x = np.array([(1, 0), (0, 1)], dtype=[('x', '<i4'), ('y', '<i4')])
    >>> x
    array([(1, 0), (0, 1)],
          dtype=[('x', '<i4'), ('y', '<i4')])
    
    >>> np.argsort(x, order=('x','y'))
    array([1, 0])
    
    >>> np.argsort(x, order=('y','x'))
    array([0, 1])