Fix failure with Numpy 1.6

I realized that when installing scipy, numpy was always getting updated to 1.8 so I fixed that in 43361afb1a965c836e9e74abf2b06a293a0f2a39. Now there is a failure with Numpy 1.6:
=================================== FAILURES ===================================
_______________________ test_mask_to_mirrored_num_range ________________________
    def test_mask_to_mirrored_num_range():
        """
        Test mask_to_mirrored_num when mirrored pixels are outside of the
        image.
        """
        center = (2.5, 2.5)
        data = np.arange(16).reshape(4, 4)
        mask = np.zeros_like(data, dtype=bool)
        mask[0, 0] = True
        mask[1, 1] = True
        data_ref = data.copy()
        data_ref[0, 0] = 0.
        data_ref[1, 1] = 0.
>       mirror_data = mask_to_mirrored_num(data, mask, center)
photutils/extern/imageutils/tests/test_array_utils.py:92: 
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ 
image = array([[ 0,  1,  2,  3],
       [ 4,  5,  6,  7],
       [ 8,  9, 10, 11],
       [12, 13, 14, 15]])
mask_image = array([[ True, False, False, False],
       [False,  True, False, False],
       [False, False, False, False],
       [False, False, False, False]], dtype=bool)
center_position = (2.5, 2.5), bbox = [0, 4, 0, 4]
    def mask_to_mirrored_num(image, mask_image, center_position, bbox=None):
        """
        Replace masked pixels with the value of the pixel mirrored across a
        given ``center_position``.  If the mirror pixel is unavailable (i.e.
        itself masked or outside of the image), then the masked pixel value
        is set to zero.

        Parameters
        ----------
        image : `numpy.ndarray`, 2D
            The 2D array of the image.

        mask_image : array-like, bool
            A boolean mask with the same shape as ``image``, where a `True`
            value indicates the corresponding element of ``image`` is
            considered bad.

        center_position : 2-tuple
            (x, y) center coordinates around which masked pixels will be
            mirrored.

        bbox : list, tuple, `numpy.ndarray`, optional
            The bounding box (x_min, x_max, y_min, y_max) over which to
            replace masked pixels.

        Returns
        -------
        result : `numpy.ndarray`, 2D
            A 2D array with replaced masked pixels.

        Examples
        --------
        >>> import numpy as np
        >>> from imageutils import mask_to_mirrored_num
        >>> image = np.arange(16).reshape(4, 4)
        >>> mask = np.zeros_like(image, dtype=bool)
        >>> mask[0, 0] = True
        >>> mask[1, 1] = True
        >>> mask_to_mirrored_num(image, mask, (1.5, 1.5))
        array([[15,  1,  2,  3],
               [ 4, 10,  6,  7],
               [ 8,  9, 10, 11],
               [12, 13, 14, 15]])
        """

        if bbox is None:
            ny, nx = image.shape
            bbox = [0, nx, 0, ny]
        subdata = copy.deepcopy(image[bbox[2]:bbox[3]+1, bbox[0]:bbox[1]+1])
        submask = mask_image[bbox[2]:bbox[3]+1, bbox[0]:bbox[1]+1]
        y_masked, x_masked = np.nonzero(submask)
        x_mirror = (2 * (center_position[0] - bbox[0])
                    - x_masked + 0.5).astype('int32')
        y_mirror = (2 * (center_position[1] - bbox[2])
                    - y_masked + 0.5).astype('int32')

        # Reset mirrored pixels that go out of the image.
        outofimage = ((x_mirror < 0) | (y_mirror < 0) |
                      (x_mirror >= subdata.shape[1]) |
                      (y_mirror >= subdata.shape[0]))
        if outofimage.any():
>           x_mirror[outofimage] = x_masked[outofimage]
E           TypeError: array cannot be safely cast to required type
photutils/extern/imageutils/array_utils.py:223: TypeError
=============== 1 failed, 295 passed, 2 skipped in 27.67 seconds ===============
astropy / imageutils

Fix failure with Numpy 1.6 #22
