Как читать документацию на примере модуля Numpy?

def array(p_object, dtype=None, copy=True, order=None, subok=False, ndmin=0): # real signature unknown; restored from __doc__
    """
    array(object, dtype=None, copy=True, order=None, subok=False, ndmin=0)
    
        Create an array.
    
        Parameters
        ----------
        object : array_like
            An array, any object exposing the array interface, an
            object whose __array__ method returns an array, or any
            (nested) sequence.
        dtype : data-type, optional
            The desired data-type for the array.  If not given, then
            the type will be determined as the minimum type required
            to hold the objects in the sequence.  This argument can only
            be used to 'upcast' the array.  For downcasting, use the
            .astype(t) method.
        copy : bool, optional
            If true (default), then the object is copied.  Otherwise, a copy
            will only be made if __array__ returns a copy, if obj is a
            nested sequence, or if a copy is needed to satisfy any of the other
            requirements (`dtype`, `order`, etc.).
        order : {'C', 'F', 'A'}, optional
            Specify the order of the array.  If order is 'C' (default), then the
            array will be in C-contiguous order (last-index varies the
            fastest).  If order is 'F', then the returned array
            will be in Fortran-contiguous order (first-index varies the
            fastest).  If order is 'A', then the returned array may
            be in any order (either C-, Fortran-contiguous, or even
            discontiguous).
        subok : bool, optional
            If True, then sub-classes will be passed-through, otherwise
            the returned array will be forced to be a base-class array (default).
        ndmin : int, optional
            Specifies the minimum number of dimensions that the resulting
            array should have.  Ones will be pre-pended to the shape as
            needed to meet this requirement.
    
        Returns
        -------
        out : ndarray
            An array object satisfying the specified requirements.
    
        See Also
        --------
        empty, empty_like, zeros, zeros_like, ones, ones_like, fill
    
        Examples
        --------
        >>> np.array([1, 2, 3])
        array([1, 2, 3])
    
        Upcasting:
    
        >>> np.array([1, 2, 3.0])
        array([ 1.,  2.,  3.])
    
        More than one dimension:
    
        >>> np.array([[1, 2], [3, 4]])
        array([[1, 2],
               [3, 4]])
    
        Minimum dimensions 2:
    
        >>> np.array([1, 2, 3], ndmin=2)
        array([[1, 2, 3]])
    
        Type provided:
    
        >>> np.array([1, 2, 3], dtype=complex)
        array([ 1.+0.j,  2.+0.j,  3.+0.j])
    
        Data-type consisting of more than one element:
    
        >>> x = np.array([(1,2),(3,4)],dtype=[('a','<i4'),('b','<i4')])
        >>> x['a']
        array([1, 3])
    
        Creating an array from sub-classes:
    
        >>> np.array(np.mat('1 2; 3 4'))
        array([[1, 2],
               [3, 4]])
    
        >>> np.array(np.mat('1 2; 3 4'), subok=True)
        matrix([[1, 2],
                [3, 4]])
    """
    pass