Python numpy 模块，right_shift() 实例源码

def test_NotImplemented_not_returned(self):
        # See gh-5964 and gh-2091. Some of these functions are not operator
        # related and were fixed for other reasons in the past.
        binary_funcs = [
            np.power, np.add, np.subtract, np.multiply, np.divide,
            np.true_divide, np.floor_divide, np.bitwise_and, np.bitwise_or,
            np.bitwise_xor, np.left_shift, np.right_shift, np.fmax,
            np.fmin, np.fmod, np.hypot, np.logaddexp, np.logaddexp2,
            np.logical_and, np.logical_or, np.logical_xor, np.maximum,
            np.minimum, np.mod
            ]

        # These functions still return NotImplemented. Will be fixed in
        # future.
        # bad = [np.greater, np.greater_equal, np.less, np.less_equal, np.not_equal]

        a = np.array('1')
        b = 1
        for f in binary_funcs:
            assert_raises(TypeError, f, a, b)

def test_NotImplemented_not_returned(self):
        # See gh-5964 and gh-2091. Some of these functions are not operator
        # related and were fixed for other reasons in the past.
        binary_funcs = [
            np.power, np.add, np.subtract, np.multiply, np.divide,
            np.true_divide, np.floor_divide, np.bitwise_and, np.bitwise_or,
            np.bitwise_xor, np.left_shift, np.right_shift, np.fmax,
            np.fmin, np.fmod, np.hypot, np.logaddexp, np.logaddexp2,
            np.logical_and, np.logical_or, np.logical_xor, np.maximum,
            np.minimum, np.mod
            ]

        # These functions still return NotImplemented. Will be fixed in
        # future.
        # bad = [np.greater, np.greater_equal, np.less, np.less_equal, np.not_equal]

        a = np.array('1')
        b = 1
        for f in binary_funcs:
            assert_raises(TypeError, f, a, b)

def test_NotImplemented_not_returned(self):
        # See gh-5964 and gh-2091. Some of these functions are not operator
        # related and were fixed for other reasons in the past.
        binary_funcs = [
            np.power, np.add, np.subtract, np.multiply, np.divide,
            np.true_divide, np.floor_divide, np.bitwise_and, np.bitwise_or,
            np.bitwise_xor, np.left_shift, np.right_shift, np.fmax,
            np.fmin, np.fmod, np.hypot, np.logaddexp, np.logaddexp2,
            np.logical_and, np.logical_or, np.logical_xor, np.maximum,
            np.minimum, np.mod
            ]

        # These functions still return NotImplemented. Will be fixed in
        # future.
        # bad = [np.greater, np.greater_equal, np.less, np.less_equal, np.not_equal]

        a = np.array('1')
        b = 1
        for f in binary_funcs:
            assert_raises(TypeError, f, a, b)

def test_NotImplemented_not_returned(self):
        # See gh-5964 and gh-2091. Some of these functions are not operator
        # related and were fixed for other reasons in the past.
        binary_funcs = [
            np.power, np.add, np.subtract, np.multiply, np.divide,
            np.true_divide, np.floor_divide, np.bitwise_and, np.bitwise_or,
            np.bitwise_xor, np.left_shift, np.right_shift, np.fmax,
            np.fmin, np.fmod, np.hypot, np.logaddexp, np.logaddexp2,
            np.logical_and, np.logical_or, np.logical_xor, np.maximum,
            np.minimum, np.mod
            ]

        # These functions still return NotImplemented. Will be fixed in
        # future.
        # bad = [np.greater, np.greater_equal, np.less, np.less_equal, np.not_equal]

        a = np.array('1')
        b = 1
        for f in binary_funcs:
            assert_raises(TypeError, f, a, b)

def dec10216(data):
    """Unpacking the 10 bit data to 16 bit"""

    arr10 = data.astype(np.uint16).flat
    new_shape = list(data.shape[:-1]) + [(data.shape[-1] * 8) / 10]
    new_shape = [int(s) for s in new_shape]
    arr16 = np.zeros(new_shape, dtype=np.uint16)
    arr16.flat[::4] = np.left_shift(arr10[::5], 2) + \
        np.right_shift((arr10[1::5]), 6)
    arr16.flat[1::4] = np.left_shift((arr10[1::5] & 63), 4) + \
        np.right_shift((arr10[2::5]), 4)
    arr16.flat[2::4] = np.left_shift(arr10[2::5] & 15, 6) + \
        np.right_shift((arr10[3::5]), 2)
    arr16.flat[3::4] = np.left_shift(arr10[3::5] & 3, 8) + \
        arr10[4::5]
    return arr16

def dec10216(inbuf):
    arr10 = inbuf.astype(np.uint16)
    arr16 = np.zeros((int(len(arr10) * 4 / 5),), dtype=np.uint16)
    arr10_len = int((len(arr16) * 5) / 4)
    arr10 = arr10[:arr10_len]  # adjust size
    """
    /*
     * pack 4 10-bit words in 5 bytes into 4 16-bit words
     *
     * 0       1       2       3       4       5
     * 01234567890123456789012345678901234567890
     * 0         1         2         3         4
     */
    ip = &in_buffer[i];
    op = &out_buffer[j];
    op[0] = ip[0]*4 + ip[1]/64;
    op[1] = (ip[1] & 0x3F)*16 + ip[2]/16;
    op[2] = (ip[2] & 0x0F)*64 + ip[3]/4;
    op[3] = (ip[3] & 0x03)*256 +ip[4];
    """
    arr16.flat[::4] = np.left_shift(arr10[::5], 2) + \
        np.right_shift((arr10[1::5]), 6)
    arr16.flat[1::4] = np.left_shift((arr10[1::5] & 63), 4) + \
        np.right_shift((arr10[2::5]), 4)
    arr16.flat[2::4] = np.left_shift(arr10[2::5] & 15, 6) + \
        np.right_shift((arr10[3::5]), 2)
    arr16.flat[3::4] = np.left_shift(arr10[3::5] & 3, 8) + \
        arr10[4::5]
    return arr16

def test_NotImplemented_not_returned(self):
        # See gh-5964 and gh-2091. Some of these functions are not operator
        # related and were fixed for other reasons in the past.
        binary_funcs = [
            np.power, np.add, np.subtract, np.multiply, np.divide,
            np.true_divide, np.floor_divide, np.bitwise_and, np.bitwise_or,
            np.bitwise_xor, np.left_shift, np.right_shift, np.fmax,
            np.fmin, np.fmod, np.hypot, np.logaddexp, np.logaddexp2,
            np.logical_and, np.logical_or, np.logical_xor, np.maximum,
            np.minimum, np.mod
            ]

        # These functions still return NotImplemented. Will be fixed in
        # future.
        # bad = [np.greater, np.greater_equal, np.less, np.less_equal, np.not_equal]

        a = np.array('1')
        b = 1
        for f in binary_funcs:
            assert_raises(TypeError, f, a, b)

def flatten_and_pack(img, bits):
    """
    Packs reduced bit depth images into bytes and returns a flattened array
        Args:
            img (uint8 numpy array): grayscale or multi-channel image
            bits (int): 1, 2, 4, or 8 bits per channel
        Returns:
            uint8 numpy array: flattened and packed array
    """

    # pad the image at the end of the rows, so that each row ends on a byte boundary
    pixels_per_byte = 8 // bits
    if len(img.shape) > 1:
        if img.shape[1] % pixels_per_byte != 0:
            img = np.hstack((img, np.zeros((img.shape[0], pixels_per_byte - img.shape[1] % pixels_per_byte), dtype=np.uint8)))

    a = np.right_shift(img, 8-bits)                                             # reduce bit depth
    b = a.flatten()                                                             # flatten
    c = np.zeros(b.size // pixels_per_byte, dtype=np.uint8)
    for i in range(0, pixels_per_byte):
        c += np.left_shift(b[i::pixels_per_byte], (pixels_per_byte-1-i)*bits)   # pack pixels and add to result

    return c

def test_NotImplemented_not_returned(self):
        # See gh-5964 and gh-2091. Some of these functions are not operator
        # related and were fixed for other reasons in the past.
        binary_funcs = [
            np.power, np.add, np.subtract, np.multiply, np.divide,
            np.true_divide, np.floor_divide, np.bitwise_and, np.bitwise_or,
            np.bitwise_xor, np.left_shift, np.right_shift, np.fmax,
            np.fmin, np.fmod, np.hypot, np.logaddexp, np.logaddexp2,
            np.logical_and, np.logical_or, np.logical_xor, np.maximum,
            np.minimum, np.mod
            ]

        # These functions still return NotImplemented. Will be fixed in
        # future.
        # bad = [np.greater, np.greater_equal, np.less, np.less_equal, np.not_equal]

        a = np.array('1')
        b = 1
        for f in binary_funcs:
            assert_raises(TypeError, f, a, b)

def test_NotImplemented_not_returned(self):
        # See gh-5964 and gh-2091. Some of these functions are not operator
        # related and were fixed for other reasons in the past.
        binary_funcs = [
            np.power, np.add, np.subtract, np.multiply, np.divide,
            np.true_divide, np.floor_divide, np.bitwise_and, np.bitwise_or,
            np.bitwise_xor, np.left_shift, np.right_shift, np.fmax,
            np.fmin, np.fmod, np.hypot, np.logaddexp, np.logaddexp2,
            np.logical_and, np.logical_or, np.logical_xor, np.maximum,
            np.minimum, np.mod
            ]

        # These functions still return NotImplemented. Will be fixed in
        # future.
        # bad = [np.greater, np.greater_equal, np.less, np.less_equal, np.not_equal]

        a = np.array('1')
        b = 1
        for f in binary_funcs:
            assert_raises(TypeError, f, a, b)

def _rgb_integers_to_components(self, rgb_integers):
        red_mask = 0x00FF0000
        green_mask = 0x0000FF00
        blue_mask =  0x000000FF
        masks = np.asarray([[red_mask, green_mask, blue_mask]])
        masked_rgb_components = np.bitwise_and(rgb_integers, masks)

        red_shifted = np.right_shift(masked_rgb_components[:,0], 16)
        green_shifted = np.right_shift(masked_rgb_components[:,1], 8)
        blue_shifted =  np.right_shift(masked_rgb_components[:,2], 0)
        return np.array([red_shifted, green_shifted, blue_shifted]).transpose()

def __rshift__(self, other):
        return right_shift(self, other)

def __irshift__(self, other):
        return right_shift(self, other, self)

def __rrshift__(self, other):
        return right_shift(other, self)

def doDecompression( packedDataList, shape, n_threads ):
    blosc.set_nthreads( n_threads )
    dataList = [None] * len(packedDataList)
    for J in np.arange(len(packedDataList) ):
#        dataStack[J,:,:] = np.reshape( 
#            np.frombuffer( blosc.decompress( packedDataList[J] ), dtype='uint8' ),
#            shape[1:] )
        # Something here Numpy-side is very slow, so let's not include that in our 
        # benchmark.
        dataList[J] = blosc.decompress( packedDataList[J] )
    return dataList




#t_half0 = time.time()
#halfimage = dm4image_8bit[:,:,::2] + np.left_shift(dm4image_8bit[:,:,1::2],4)
#t_half1 = time.time()
#restoreimage = np.empty( header['dimensions'], dtype='uint8' )
##image[0::2] = np.left_shift(interlaced_image,4)/16
##image[1::2] = np.right_shift(interlaced_image,4)
## Different interlace option
## TODO: AND array with 15 instead?
#restoreimage[:,:,::2] = (np.left_shift( halfimage, 4 ) & 15 )
#restoreimage[:,:,1::2] = np.right_shift( halfimage, 4 )
#t_half2 = time.time()
#
#print( "4-byte encoding time (s): %f" % (t_half1 - t_half0) )
#print( "4-byte DEcoding time (s): %f" % (t_half2 - t_half1) )

def numpy_int16_to_uint8(depth):
        x, y = depth.shape
        out = np.ndarray((x, y, 2), dtype=np.uint8)
        out[:, :, 0] = np.right_shift(depth, 8)
        out[:, :, 1] = depth.astype(np.uint8)
        return out

def binary_permutations(n_bits):
    """
    Given some number of bits, return a shape (2**n_bits, n_bits) boolean array containing every permoutation
    of those bits as a row.
    :param n_bits: An integer number of bits
    :return: A shape (2**n_bits, n_bits) boolean array containing every permoutation
        of those bits as a row.
    """
    return np.right_shift(np.arange(2**n_bits)[:, None], np.arange(n_bits-1, -1, -1)[None, :]) & 1

def _mix(a, b, f):
    #Used for the blend function (mixes colors according to their alpha values)
    c = numpy.multiply(numpy.subtract(b,a),f)
    return numpy.add(numpy.right_shift(c,8),a)

def complex_data(self):
    '''
    This will cast each byte to an int8 and interpret each byte
    as 4 bits real values and 4 bits imag values (RRRRIIII). The data are then
    used to create a 3D numpy array of dtype=complex, which is returned. 

    The shape of the numpy array is N half frames, M subbands, K data points per half frame,
    where K = constants.bins_per_half_frame, N is typically 129 and M is typically 1 for
    compamp files and 16 for archive-compamp files. 

    Note that this returns a Numpy array of type complex64. This data is not retained within Compamp objects.

    '''
    #note that since we can only pack into int8 types, we must pad each 4-bit value with 4, 0 bits
    #this effectively multiplies each 4-bit value by 16 when that value is represented as an 8-bit signed integer.
    packed_data = self._packed_data()
    header = self.header()

    real_val = np.bitwise_and(packed_data, 0xf0).astype(np.int8)  # coef's are: RRRRIIII (4 bits real,
    imag_val = np.left_shift(np.bitwise_and(packed_data, 0x0f), 4).astype(np.int8)  # 4 bits imaginary in 2's complement)

    cdata = np.empty(len(real_val), np.complex64)

    #"Normalize" by making appropriate bit-shift. Otherwise, values for real and imaginary coefficients are
    #inflated by 16x. 
    cdata.real = np.right_shift(real_val, 4)
    cdata.imag = np.right_shift(imag_val, 4)

    # expose compamp measurement blocks
    cdata = cdata.reshape((header['number_of_half_frames'], header['number_of_subbands'], constants.bins_per_half_frame))

    return cdata

def _get_voc_color_map(n=256):
    color_map = np.zeros((n, 3))
    for i in xrange(n):
        r = b = g = 0
        cid = i
        for j in xrange(0, 8):
            r = np.bitwise_or(r, np.left_shift(np.unpackbits(np.array([cid], dtype=np.uint8))[-1], 7-j))
            g = np.bitwise_or(g, np.left_shift(np.unpackbits(np.array([cid], dtype=np.uint8))[-2], 7-j))
            b = np.bitwise_or(b, np.left_shift(np.unpackbits(np.array([cid], dtype=np.uint8))[-3], 7-j))
            cid = np.right_shift(cid, 3)

        color_map[i][0] = r
        color_map[i][1] = g
        color_map[i][2] = b
    return color_map