def average(a, axis=None, weights=None, returned=False): # pylint: disable=missing-docstring
if axis is not None and not isinstance(axis, six.integer_types):
# TODO(wangpeng): Support tuple of ints as `axis`
raise ValueError('`axis` must be an integer. Tuple of ints is not '
'supported yet. Got type: %s' % type(axis))
a = array_ops.array(a)
if weights is None: # Treat all weights as 1
if not np.issubdtype(a.dtype, np.inexact):
a = a.astype(
utils.result_type(a.dtype, dtypes.default_float_type()))
avg = tf.reduce_mean(a.data, axis=axis)
if returned:
if axis is None:
weights_sum = tf.size(a.data)
else:
weights_sum = tf.shape(a.data)[axis]
weights_sum = tf.cast(weights_sum, a.data.dtype)
else:
if np.issubdtype(a.dtype, np.inexact):
out_dtype = utils.result_type(a.dtype, weights)
else:
out_dtype = utils.result_type(a.dtype, weights,
dtypes.default_float_type())
a = array_ops.array(a, out_dtype).data
weights = array_ops.array(weights, out_dtype).data
def rank_equal_case():
tf.debugging.Assert(
tf.reduce_all(tf.shape(a) == tf.shape(weights)),
[tf.shape(a), tf.shape(weights)])
weights_sum = tf.reduce_sum(weights, axis=axis)
avg = tf.reduce_sum(a * weights, axis=axis) / weights_sum
return avg, weights_sum
if axis is None:
avg, weights_sum = rank_equal_case()
else:
def rank_not_equal_case():
tf.debugging.Assert(tf.rank(weights) == 1, [tf.rank(weights)])
weights_sum = tf.reduce_sum(weights)
axes = tf.convert_to_tensor([[axis], [0]])
avg = tf.tensordot(a, weights, axes) / weights_sum
return avg, weights_sum
# We condition on rank rather than shape equality, because if we do the
# latter, when the shapes are partially unknown but the ranks are known
# and different, utils.cond will run shape checking on the true branch,
# which will raise a shape-checking error.
avg, weights_sum = utils.cond(
tf.rank(a) == tf.rank(weights), rank_equal_case,
rank_not_equal_case)
avg = array_ops.array(avg)
if returned:
weights_sum = array_ops.broadcast_to(weights_sum, tf.shape(avg.data))
return avg, weights_sum
return avg
def array(val, dtype=None, copy=True, ndmin=0): # pylint: disable=redefined-outer-name
"""Creates an ndarray with the contents of val.
Args:
val: array_like. Could be an ndarray, a Tensor or any object that can be
converted to a Tensor using `tf.convert_to_tensor`.
dtype: Optional, defaults to dtype of the `val`. The type of the resulting
ndarray. Could be a python type, a NumPy type or a TensorFlow `DType`.
copy: Determines whether to create a copy of the backing buffer. Since
Tensors are immutable, a copy is made only if val is placed on a different
device than the current one. Even if `copy` is False, a new Tensor may
need to be built to satisfy `dtype` and `ndim`. This is used only if `val`
is an ndarray or a Tensor.
ndmin: The minimum rank of the returned array.
Returns:
An ndarray.
"""
if dtype:
dtype = utils.result_type(dtype)
if isinstance(val, arrays_lib.ndarray):
result_t = val.data
else:
result_t = val
if copy and isinstance(result_t, tf.Tensor):
# Note: In eager mode, a copy of `result_t` is made only if it is not on
# the context device.
result_t = tf.identity(result_t)
if not isinstance(result_t, tf.Tensor):
if not dtype:
dtype = utils.result_type(result_t)
# We can't call `convert_to_tensor(result_t, dtype=dtype)` here because
# convert_to_tensor doesn't allow incompatible arguments such as (5.5, int)
# while np.array allows them. We need to convert-then-cast.
def maybe_data(x):
if isinstance(x, arrays_lib.ndarray):
return x.data
return x
# Handles lists of ndarrays
result_t = tf.nest.map_structure(maybe_data, result_t)
result_t = arrays_lib.convert_to_tensor(result_t)
result_t = tf.cast(result_t, dtype=dtype)
elif dtype:
result_t = tf.cast(result_t, dtype)
ndims = tf.rank(result_t)
def true_fn():
old_shape = tf.shape(result_t)
new_shape = tf.concat([tf.ones(ndmin - ndims, tf.int32), old_shape],
axis=0)
return tf.reshape(result_t, new_shape)
result_t = utils.cond(utils.greater(ndmin, ndims), true_fn,
lambda: result_t)
return arrays_lib.tensor_to_ndarray(result_t)
def eye(N, M=None, k=0, dtype=float): # pylint: disable=invalid-name,missing-docstring
if dtype:
dtype = utils.result_type(dtype)
if not M:
M = N
# Making sure N, M and k are `int`
N = int(N)
M = int(M)
k = int(k)
if k >= M or -k >= N:
# tf.linalg.diag will raise an error in this case
return zeros([N, M], dtype=dtype)
if k == 0:
return arrays_lib.tensor_to_ndarray(tf.eye(N, M, dtype=dtype))
# We need the precise length, otherwise tf.linalg.diag will raise an error
diag_len = min(N, M)
if k > 0:
if N >= M:
diag_len -= k
elif N + k > M:
diag_len = M - k
elif k <= 0:
if M >= N:
diag_len += k
elif M - k > N:
diag_len = N + k
diagonal = tf.ones([diag_len], dtype=dtype)
return arrays_lib.tensor_to_ndarray(
tf.linalg.diag(diagonal=diagonal, num_rows=N, num_cols=M, k=k))
def linspace(start, stop, num=50, endpoint=True, retstep=False, dtype=float):
if dtype:
dtype = utils.result_type(dtype)
start = array_creation.asarray(start, dtype=dtype)
stop = array_creation.asarray(stop, dtype=dtype)
if num == 0:
return empty(dtype)
if num < 0:
raise ValueError(
'Number of samples {} must be non-negative.'.format(num))
step = np.nan
if endpoint:
result = tf.linspace(start.data, stop.data, num)
if num > 1:
step = (stop - start) / (num - 1)
else:
# tf.linspace does not support endpoint=False so we manually handle it
# here.
if num > 1:
step = (stop - start) / num
result = tf.linspace(start.data, (stop - step).data, num)
else:
result = tf.linspace(start.data, stop.data, num)
if dtype:
result = tf.cast(result, dtype)
if retstep:
return arrays.tensor_to_ndarray(result), step
else:
return arrays.tensor_to_ndarray(result)
def array(val, dtype=None, copy=True, ndmin=0):
"""Creates an ndarray with the contents of val.
Args:
val: array_like. Could be an ndarray, a Tensor or any object that can
be converted to a Tensor using `tf.convert_to_tensor`.
dtype: Optional, defaults to dtype of the `val`. The type of the
resulting ndarray. Could be a python type, a NumPy type or a TensorFlow
`DType`.
copy: Determines whether to create a copy of the backing buffer. Since
Tensors are immutable, a copy is made only if val is placed on a different
device than the current one. Even if `copy` is False, a new Tensor may
need to be built to satisfy `dtype` and `ndim`. This is used only if `val`
is an ndarray or a Tensor.
ndmin: The minimum rank of the returned array.
Returns:
An ndarray.
"""
if dtype:
dtype = utils.result_type(dtype)
if isinstance(val, arrays.ndarray):
result_t = val.data
else:
result_t = val
if copy and isinstance(result_t, tf.Tensor):
# Note: In eager mode, a copy of `result_t` is made only if it is not on
# the context device.
result_t = tf.identity(result_t)
if not isinstance(result_t, tf.Tensor):
if not dtype:
dtype = utils.result_type(result_t)
# We can't call `convert_to_tensor(result_t, dtype=dtype)` here because
# convert_to_tensor doesn't allow incompatible arguments such as (5.5, int)
# while np.array allows them. We need to convert-then-cast.
result_t = arrays.convert_to_tensor(result_t)
result_t = tf.cast(result_t, dtype=dtype)
elif dtype:
result_t = tf.cast(result_t, dtype)
ndims = len(result_t.shape)
if ndmin > ndims:
old_shape = list(result_t.shape)
new_shape = [1 for _ in range(ndmin - ndims)] + old_shape
result_t = tf.reshape(result_t, new_shape)
return arrays.tensor_to_ndarray(result_t)
def concatenate(arys, axis=0):
if not isinstance(arys, (list, tuple)):
arys = [arys]
if not arys:
raise ValueError('Need at least one array to concatenate.')
dtype = utils.result_type(*arys)
arys = [array_ops.array(array, dtype=dtype).data for array in arys]
return arrays.tensor_to_ndarray(tf.concat(arys, axis))
def logspace(start, stop, num=50, endpoint=True, base=10.0, dtype=None):
if dtype:
dtype = utils.result_type(dtype)
result = linspace(start, stop, num=num, endpoint=endpoint)
result = tf.pow(base, result.data)
if dtype:
result = tf.cast(result, dtype)
return arrays.tensor_to_ndarray(result)
def _reduce(tf_fn, a, axis=None, dtype=None, keepdims=None,
promote_int=_TO_INT64, tf_bool_fn=None, preserve_bool=False):
"""A general reduction function.
Args:
tf_fn: the TF reduction function.
a: the array to be reduced.
axis: (optional) the axis along which to do the reduction. If None, all
dimensions are reduced.
dtype: (optional) the dtype of the result.
keepdims: (optional) whether to keep the reduced dimension(s).
promote_int: how to promote integer and bool inputs. There are three
choices: (1) _TO_INT64: always promote them to int64 or uint64; (2)
_TO_FLOAT: always promote them to a float type (determined by
dtypes.default_float_type); (3) None: don't promote.
tf_bool_fn: (optional) the TF reduction function for bool inputs. It
will only be used if `dtype` is explicitly set to `np.bool_` or if `a`'s
dtype is `np.bool_` and `preserve_bool` is True.
preserve_bool: a flag to control whether to use `tf_bool_fn` if `a`'s dtype
is `np.bool_` (some reductions such as np.sum convert bools to
integers, while others such as np.max preserve bools.
Returns:
An ndarray.
"""
if dtype:
dtype = utils.result_type(dtype)
if keepdims is None:
keepdims = False
a = array_creation.asarray(a, dtype=dtype)
if ((dtype == np.bool_ or preserve_bool and a.dtype == np.bool_)
and tf_bool_fn is not None):
return utils.tensor_to_ndarray(
tf_bool_fn(input_tensor=a.data, axis=axis, keepdims=keepdims))
if dtype is None:
dtype = a.dtype
if np.issubdtype(dtype, np.integer) or dtype == np.bool_:
if promote_int == _TO_INT64:
# If a is an integer/bool type and whose bit width is less than 64,
# numpy up-casts it to 64-bit.
if dtype == np.bool_:
is_signed = True
width = 8 # We can use any number here that is less than 64
else:
is_signed = np.issubdtype(dtype, np.signedinteger)
width = np.iinfo(dtype).bits
if width < 64:
if is_signed:
dtype = np.int64
else:
dtype = np.uint64
a = a.astype(dtype)
elif promote_int == _TO_FLOAT:
a = a.astype(dtypes.default_float_type())
return utils.tensor_to_ndarray(
tf_fn(input_tensor=a.data, axis=axis, keepdims=keepdims))
def promote_args_types(a, b):
a = asarray(a)
b = asarray(b)
output_type = utils.result_type(a.dtype, b.dtype)
if output_type != a.dtype:
a = a.astype(output_type)
if output_type != b.dtype:
b = b.astype(output_type)
return (a, b)
def _comparison(tf_fun, x1, x2, cast_bool_to_int=False):
dtype = utils.result_type(x1, x2)
# Cast x1 and x2 to the result_type if needed.
x1 = array_creation.asarray(x1, dtype=dtype)
x2 = array_creation.asarray(x2, dtype=dtype)
x1 = x1.data
x2 = x2.data
if cast_bool_to_int and x1.dtype == tf.bool:
x1 = tf.cast(x1, tf.int32)
x2 = tf.cast(x2, tf.int32)
return utils.tensor_to_ndarray(tf_fun(x1, x2))
def ones_like(a, dtype=None):
"""Returns an array of ones with the shape and type of the input array.
Args:
a: array_like. Could be an ndarray, a Tensor or any object that can be
converted to a Tensor using `tf.convert_to_tensor`.
dtype: Optional, defaults to dtype of the input array. The type of the
resulting ndarray. Could be a python type, a NumPy type or a TensorFlow
`DType`.
Returns:
An ndarray.
"""
if isinstance(a, arrays_lib.ndarray):
a = a.data
if dtype is None:
dtype = utils.result_type(a)
else:
dtype = utils.result_type(dtype)
return arrays_lib.tensor_to_ndarray(tf.ones_like(a, dtype))
def arange(start, stop=None, step=1, dtype=None):
"""Returns `step`-separated values in the range [start, stop).
Args:
start: Start of the interval. Included in the range.
stop: End of the interval. If not specified, `start` is treated as 0 and
`start` value is used as `stop`. If specified, it is not included in the
range if `step` is integer. When `step` is floating point, it may or may
not be included.
step: The difference between 2 consecutive values in the output range.
It is recommended to use `linspace` instead of using non-integer values
for `step`.
dtype: Optional. Type of the resulting ndarray. Could be a python type, a
NumPy type or a TensorFlow `DType`. If not provided, the largest type of
`start`, `stop`, `step` is used.
Raises:
ValueError: If step is zero.
"""
if not step:
raise ValueError('step must be non-zero.')
if dtype:
dtype = utils.to_tf_type(dtype)
else:
dtype = utils.result_type(
utils.array_dtype(start).as_numpy_dtype,
utils.array_dtype(step).as_numpy_dtype)
if stop is not None:
dtype = utils.result_type(dtype,
utils.array_dtype(stop).as_numpy_dtype)
dtype = utils.to_tf_type(dtype)
if step > 0 and ((stop is not None and start > stop) or
(stop is None and start < 0)):
return array([], dtype=dtype)
if step < 0 and ((stop is not None and start < stop) or
(stop is None and start > 0)):
return array([], dtype=dtype)
# TODO(srbs): There are some bugs when start or stop is float type and dtype
# is integer type.
return utils.tensor_to_ndarray(
tf.cast(tf.range(start, limit=stop, delta=step), dtype=dtype))
def full_like(a, fill_value, dtype=None, order='K', subok=True, shape=None): # pylint: disable=missing-docstring,redefined-outer-name
"""order, subok and shape arguments mustn't be changed."""
if order != 'K':
raise ValueError('Non-standard orders are not supported.')
if not subok:
raise ValueError('subok being False is not supported.')
if shape:
raise ValueError('Overriding the shape is not supported.')
a = asarray(a).data
dtype = dtype or utils.result_type(a)
fill_value = asarray(fill_value, dtype=dtype)
return arrays_lib.tensor_to_ndarray(
tf.broadcast_to(fill_value.data, tf.shape(a)))
def trace(a, offset=0, axis1=0, axis2=1, dtype=None): # pylint: disable=missing-docstring
if dtype:
dtype = utils.result_type(dtype)
a = array_ops.asarray(a, dtype).data
if offset == 0:
a_shape = a.shape
if a_shape.rank is not None:
rank = len(a_shape)
if (axis1 == -2 or axis1 == rank - 2) and (axis2 == -1
or axis2 == rank - 1):
return utils.tensor_to_ndarray(tf.linalg.trace(a))
a = array_ops.diagonal(a, offset, axis1, axis2)
return array_ops.sum(a, -1, dtype)
def zeros_like(a, dtype=None):
"""Returns an array of zeros with the shape and type of the input array.
Args:
a: array_like. Could be an ndarray, a Tensor or any object that can be
converted to a Tensor using `tf.convert_to_tensor`.
dtype: Optional, defaults to dtype of the input array. The type of the
resulting ndarray. Could be a python type, a NumPy type or a TensorFlow
`DType`.
Returns:
An ndarray.
"""
if isinstance(a, arrays_lib.ndarray):
a = a.data
if dtype is None:
# We need to let utils.result_type decide the dtype, not tf.zeros_like
dtype = utils.result_type(a)
else:
# TF and numpy has different interpretations of Python types such as
# `float`, so we let `utils.result_type` decide.
dtype = utils.result_type(dtype)
dtype = tf.as_dtype(dtype) # Work around b/149877262
return arrays_lib.tensor_to_ndarray(tf.zeros_like(a, dtype))
def logspace(start,
stop,
num=50,
endpoint=True,
base=10.0,
dtype=None,
axis=0):
dtype = utils.result_type(start, stop, dtype)
result = linspace(start,
stop,
num=num,
endpoint=endpoint,
dtype=dtype,
axis=axis).data
result = tf.pow(tf.cast(base, result.dtype), result)
if dtype:
result = tf.cast(result, dtype)
return arrays.tensor_to_ndarray(result)
def zeros(shape, dtype=float): # pylint: disable=redefined-outer-name
"""Returns an ndarray with the given shape and type filled with zeros.
Args:
shape: A fully defined shape. Could be - NumPy array or a python scalar,
list or tuple of integers, - TensorFlow tensor/ndarray of integer type and
rank <=1.
dtype: Optional, defaults to float. The type of the resulting ndarray. Could
be a python type, a NumPy type or a TensorFlow `DType`.
Returns:
An ndarray.
"""
if dtype:
dtype = utils.result_type(dtype)
if isinstance(shape, arrays_lib.ndarray):
shape = shape.data
return arrays_lib.tensor_to_ndarray(tf.zeros(shape, dtype=dtype))
def geomspace(start, stop, num=50, endpoint=True, dtype=float): # pylint: disable=missing-docstring
if dtype:
dtype = utils.result_type(dtype)
if num < 0:
raise ValueError('Number of samples {} must be non-negative.'.format(num))
if not num:
return empty([0])
step = 1.
if endpoint:
if num > 1:
step = tf.pow((stop / start), 1 / (num - 1))
else:
step = tf.pow((stop / start), 1 / num)
result = tf.cast(tf.range(num), step.dtype)
result = tf.pow(step, result)
result = tf.multiply(result, start)
if dtype:
result = tf.cast(result, dtype=dtype)
return arrays_lib.tensor_to_ndarray(result)
def full_like(a, fill_value, dtype=None):
"""Returns an array with same shape and dtype as `a` filled with `fill_value`.
Args:
a: array_like. Could be an ndarray, a Tensor or any object that can be
converted to a Tensor using `tf.convert_to_tensor`.
fill_value: array_like. Could be an ndarray, a Tensor or any object that can
be converted to a Tensor using `tf.convert_to_tensor`.
dtype: Optional, defaults to dtype of the `a`. The type of the resulting
ndarray. Could be a python type, a NumPy type or a TensorFlow `DType`.
Returns:
An ndarray.
Raises:
ValueError: if `fill_value` can not be broadcast to shape `shape`.
"""
a = asarray(a)
dtype = dtype or utils.result_type(a)
return full(a.shape, fill_value, dtype)
def equal(x1, x2):
"""Compare two arrays for equality element-wise.
Both arrays must either be of the same shape or one should be broadcastable
to the other.
Args:
x1: array_like. Could be an ndarray, a Tensor or any object that can
be converted to a Tensor using `tf.convert_to_tensor`.
x2: array_like. Could be an ndarray, a Tensor or any object that can
be converted to a Tensor using `tf.convert_to_tensor`.
Returns:
An ndarray of type bool and broadcasted shape of x1 and x2.
"""
dtype = utils.result_type(x1, x2)
# Cast x1 and x2 to the result_type if needed.
x1 = array_creation.array(x1, copy=False, dtype=dtype)
x2 = array_creation.array(x2, copy=False, dtype=dtype)
return utils.tensor_to_ndarray(tf.equal(x1.data, x2.data))
def tri(N, M=None, k=0, dtype=None): # pylint: disable=invalid-name,missing-docstring
M = M if M is not None else N
if dtype is not None:
dtype = utils.result_type(dtype)
else:
dtype = dtypes.default_float_type()
if k < 0:
lower = -k - 1
if lower > N:
r = tf.zeros([N, M], dtype)
else:
# Keep as tf bool, since we create an upper triangular matrix and invert
# it.
o = tf.ones([N, M], dtype=tf.bool)
r = tf.cast(tf.math.logical_not(tf.linalg.band_part(o, lower, -1)), dtype)
else:
o = tf.ones([N, M], dtype)
if k > M:
r = o
else:
r = tf.linalg.band_part(o, -1, k)
return utils.tensor_to_ndarray(r)
def linspace( # pylint: disable=missing-docstring
start,
stop,
num=50,
endpoint=True,
retstep=False,
dtype=float,
axis=0):
if dtype:
dtype = utils.result_type(dtype)
start = array_ops.array(start, dtype=dtype).data
stop = array_ops.array(stop, dtype=dtype).data
if num < 0:
raise ValueError(
'Number of samples {} must be non-negative.'.format(num))
step = tf.convert_to_tensor(np.nan)
if endpoint:
result = tf.linspace(start, stop, num, axis=axis)
if num > 1:
step = (stop - start) / (num - 1)
else:
# tf.linspace does not support endpoint=False so we manually handle it
# here.
if num > 1:
step = ((stop - start) / num)
new_stop = tf.cast(stop, step.dtype) - step
start = tf.cast(start, new_stop.dtype)
result = tf.linspace(start, new_stop, num, axis=axis)
else:
result = tf.linspace(start, stop, num, axis=axis)
if dtype:
result = tf.cast(result, dtype)
if retstep:
return arrays.tensor_to_ndarray(result), arrays.tensor_to_ndarray(step)
else:
return arrays.tensor_to_ndarray(result)
def asarray(a, dtype=None):
if dtype:
dtype = utils.result_type(dtype)
if isinstance(a, arrays_lib.ndarray) and (not dtype or dtype == a.dtype):
return a
return array(a, dtype, copy=False)
def concatenate(arys, axis=0):
if not arys:
raise ValueError('Need at least one array to concatenate.')
dtype = utils.result_type(*arys)
arys = [array_creation.asarray(array, dtype=dtype).data for array in arys]
return arrays.tensor_to_ndarray(tf.concat(arys, axis))
def _promote_dtype(*arrays): dtype = utils.result_type(*arrays) return [asarray(a, dtype=dtype) for a in arrays]
