提交 f40302ca 编写于 作者: A A. Unique TensorFlower 提交者: TensorFlower Gardener

Update generated Python Op docs.

Change: 125729946
上级 aafb9be5
...@@ -727,7 +727,7 @@ tf.shape(tf.concat(1, [t3, t4])) ==> [2, 6] ...@@ -727,7 +727,7 @@ tf.shape(tf.concat(1, [t3, t4])) ==> [2, 6]
- - - - - -
### `tf.pack(values, name='pack')` {#pack} ### `tf.pack(values, axis=0, name='pack')` {#pack}
Packs a list of rank-`R` tensors into one rank-`(R+1)` tensor. Packs a list of rank-`R` tensors into one rank-`(R+1)` tensor.
...@@ -743,6 +743,8 @@ This is the opposite of unpack. The numpy equivalent is ...@@ -743,6 +743,8 @@ This is the opposite of unpack. The numpy equivalent is
* <b>`values`</b>: A list of `Tensor` objects with the same shape and type. * <b>`values`</b>: A list of `Tensor` objects with the same shape and type.
* <b>`axis`</b>: An `int`. The axis to pack along. Defaults to the first dimension.
Supports negative indexes.
* <b>`name`</b>: A name for this operation (optional). * <b>`name`</b>: A name for this operation (optional).
##### Returns: ##### Returns:
...@@ -750,16 +752,21 @@ This is the opposite of unpack. The numpy equivalent is ...@@ -750,16 +752,21 @@ This is the opposite of unpack. The numpy equivalent is
* <b>`output`</b>: A packed `Tensor` with the same type as `values`. * <b>`output`</b>: A packed `Tensor` with the same type as `values`.
##### Raises:
* <b>`ValueError`</b>: If `axis` is out of the range [-(R+1), R+1).
- - - - - -
### `tf.unpack(value, num=None, name='unpack')` {#unpack} ### `tf.unpack(value, num=None, axis=0, name='unpack')` {#unpack}
Unpacks the outer dimension of a rank-`R` tensor into rank-`(R-1)` tensors. Unpacks the given dimension of a rank-`R` tensor into rank-`(R-1)` tensors.
Unpacks `num` tensors from `value` along the first dimension. Unpacks `num` tensors from `value` along the given dimension.
If `num` is not specified (the default), it is inferred from `value`'s shape. If `num` is not specified (the default), it is inferred from `value`'s shape.
If `value.shape[0]` is not known, `ValueError` is raised. If `value.shape[axis]` is not known, `ValueError` is raised.
The ith tensor in `output` is the slice `value[i, ...]`. Each tensor in The ith tensor in `output` is the slice `value[i, ...]`. Each tensor in
`output` has shape `value.shape[1:]`. `output` has shape `value.shape[1:]`.
...@@ -772,8 +779,10 @@ This is the opposite of pack. The numpy equivalent is ...@@ -772,8 +779,10 @@ This is the opposite of pack. The numpy equivalent is
* <b>`value`</b>: A rank `R > 0` `Tensor` to be unpacked. * <b>`value`</b>: A rank `R > 0` `Tensor` to be unpacked.
* <b>`num`</b>: An `int`. The first dimension of value. Automatically inferred if * <b>`num`</b>: An `int`. The length of the dimension `axis`. Automatically inferred
`None` (the default). if `None` (the default).
* <b>`axis`</b>: An `int`. The axis to unpack along. Defaults to the first
dimension. Supports negative indexes.
* <b>`name`</b>: A name for the operation (optional). * <b>`name`</b>: A name for the operation (optional).
##### Returns: ##### Returns:
...@@ -784,6 +793,7 @@ This is the opposite of pack. The numpy equivalent is ...@@ -784,6 +793,7 @@ This is the opposite of pack. The numpy equivalent is
* <b>`ValueError`</b>: If `num` is unspecified and cannot be inferred. * <b>`ValueError`</b>: If `num` is unspecified and cannot be inferred.
* <b>`ValueError`</b>: If `axis` is out of the range [-R, R).
- - - - - -
......
...@@ -255,6 +255,13 @@ Generate `n` samples. ...@@ -255,6 +255,13 @@ Generate `n` samples.
Standard deviation of the distribution. Standard deviation of the distribution.
- - -
#### `tf.contrib.distributions.BaseDistribution.strict` {#BaseDistribution.strict}
Boolean describing behavior on invalid input.
- - - - - -
#### `tf.contrib.distributions.BaseDistribution.variance(name='variance')` {#BaseDistribution.variance} #### `tf.contrib.distributions.BaseDistribution.variance(name='variance')` {#BaseDistribution.variance}
...@@ -441,6 +448,13 @@ Generate `n` samples. ...@@ -441,6 +448,13 @@ Generate `n` samples.
Standard deviation of the distribution. Standard deviation of the distribution.
- - -
#### `tf.contrib.distributions.ContinuousDistribution.strict` {#ContinuousDistribution.strict}
Boolean describing behavior on invalid input.
- - - - - -
#### `tf.contrib.distributions.ContinuousDistribution.variance(name='variance')` {#ContinuousDistribution.variance} #### `tf.contrib.distributions.ContinuousDistribution.variance(name='variance')` {#ContinuousDistribution.variance}
...@@ -613,6 +627,13 @@ Generate `n` samples. ...@@ -613,6 +627,13 @@ Generate `n` samples.
Standard deviation of the distribution. Standard deviation of the distribution.
- - -
#### `tf.contrib.distributions.DiscreteDistribution.strict` {#DiscreteDistribution.strict}
Boolean describing behavior on invalid input.
- - - - - -
#### `tf.contrib.distributions.DiscreteDistribution.variance(name='variance')` {#DiscreteDistribution.variance} #### `tf.contrib.distributions.DiscreteDistribution.variance(name='variance')` {#DiscreteDistribution.variance}
...@@ -862,6 +883,13 @@ Standard deviation of the distribution. ...@@ -862,6 +883,13 @@ Standard deviation of the distribution.
* <b>`std`</b>: `Tensor` of the same type and shape as `p`. * <b>`std`</b>: `Tensor` of the same type and shape as `p`.
- - -
#### `tf.contrib.distributions.Bernoulli.strict` {#Bernoulli.strict}
Boolean describing behavior on invalid input.
- - - - - -
#### `tf.contrib.distributions.Bernoulli.variance(name='variance')` {#Bernoulli.variance} #### `tf.contrib.distributions.Bernoulli.variance(name='variance')` {#Bernoulli.variance}
...@@ -895,19 +923,20 @@ Note, the following methods of the base class aren't implemented: ...@@ -895,19 +923,20 @@ Note, the following methods of the base class aren't implemented:
* log_cdf * log_cdf
- - - - - -
#### `tf.contrib.distributions.Categorical.__init__(logits, name='Categorical', dtype=tf.int32)` {#Categorical.__init__} #### `tf.contrib.distributions.Categorical.__init__(logits, dtype=tf.int32, strict=True, name='Categorical')` {#Categorical.__init__}
Initialize Categorical distributions using class log-probabilities. Initialize Categorical distributions using class log-probabilities.
##### Args: ##### Args:
* <b>`logits`</b>: An N-D `Tensor` representing the log probabilities of a set of * <b>`logits`</b>: An N-D `Tensor`, `N >= 1`, representing the log probabilities
Categorical distributions. The first N - 1 dimensions index into a of a set of Categorical distributions. The first `N - 1` dimensions
batch of independent distributions and the last dimension indexes index into a batch of independent distributions and the last dimension
into the classes. indexes into the classes.
* <b>`name`</b>: A name for this distribution (optional).
* <b>`dtype`</b>: The type of the event samples (default: int32). * <b>`dtype`</b>: The type of the event samples (default: int32).
* <b>`strict`</b>: Unused in this distribution.
* <b>`name`</b>: A name for this distribution (optional).
- - - - - -
...@@ -1074,6 +1103,13 @@ Sample `n` observations from the Categorical distribution. ...@@ -1074,6 +1103,13 @@ Sample `n` observations from the Categorical distribution.
Standard deviation of the distribution. Standard deviation of the distribution.
- - -
#### `tf.contrib.distributions.Categorical.strict` {#Categorical.strict}
Boolean describing behavior on invalid input.
- - - - - -
#### `tf.contrib.distributions.Categorical.variance(name='variance')` {#Categorical.variance} #### `tf.contrib.distributions.Categorical.variance(name='variance')` {#Categorical.variance}
...@@ -1090,16 +1126,26 @@ The Chi2 distribution with degrees of freedom df. ...@@ -1090,16 +1126,26 @@ The Chi2 distribution with degrees of freedom df.
The PDF of this distribution is: The PDF of this distribution is:
```pdf(x) = (x^(df/2 - 1)e^(-x/2))/(2^(k/2)Gamma(k/2)), x > 0``` ```pdf(x) = (x^(df/2 - 1)e^(-x/2))/(2^(df/2)Gamma(df/2)), x > 0```
Note that the Chi2 distribution is a special case of the Gamma distribution, Note that the Chi2 distribution is a special case of the Gamma distribution,
with Chi2(df) = Gamma(df/2, 1/2). with Chi2(df) = Gamma(df/2, 1/2).
- - - - - -
#### `tf.contrib.distributions.Chi2.__init__(df, name='Chi2')` {#Chi2.__init__} #### `tf.contrib.distributions.Chi2.__init__(df, strict=True, name='Chi2')` {#Chi2.__init__}
Construct Chi2 distributions with parameter `df`.
##### Args:
* <b>`df`</b>: `float` or `double` tensor, the degrees of freedom of the
distribution(s). `df` must contain only positive values.
* <b>`strict`</b>: Whether to assert that `df > 0`, and that `x > 0` in the
methods `pdf(x)` and `log_pdf(x)`. If `strict` is False
and the inputs are invalid, correct behavior is not guaranteed.
* <b>`name`</b>: The name to prepend to all ops created by this distribution.
- - - - - -
...@@ -1363,6 +1409,13 @@ See the doc for tf.random_gamma for further detail. ...@@ -1363,6 +1409,13 @@ See the doc for tf.random_gamma for further detail.
Standard deviation of this distribution. Standard deviation of this distribution.
- - -
#### `tf.contrib.distributions.Chi2.strict` {#Chi2.strict}
Boolean describing behavior on invalid input.
- - - - - -
#### `tf.contrib.distributions.Chi2.variance(name='variance')` {#Chi2.variance} #### `tf.contrib.distributions.Chi2.variance(name='variance')` {#Chi2.variance}
...@@ -1385,10 +1438,20 @@ Note that the Exponential distribution is a special case of the Gamma ...@@ -1385,10 +1438,20 @@ Note that the Exponential distribution is a special case of the Gamma
distribution, with Exponential(lam) = Gamma(1, lam). distribution, with Exponential(lam) = Gamma(1, lam).
- - - - - -
#### `tf.contrib.distributions.Exponential.__init__(lam, name='Exponential')` {#Exponential.__init__} #### `tf.contrib.distributions.Exponential.__init__(lam, strict=True, name='Exponential')` {#Exponential.__init__}
Construct Exponential distribution with parameter `lam`.
##### Args:
* <b>`lam`</b>: `float` or `double` tensor, the rate of the distribution(s).
`lam` must contain only positive values.
* <b>`strict`</b>: Whether to assert that `lam > 0`, and that `x > 0` in the
methods `pdf(x)` and `log_pdf(x)`. If `strict` is False
and the inputs are invalid, correct behavior is not guaranteed.
* <b>`name`</b>: The name to prepend to all ops created by this distribution.
- - - - - -
...@@ -1649,6 +1712,13 @@ Sample `n` observations from the Exponential Distributions. ...@@ -1649,6 +1712,13 @@ Sample `n` observations from the Exponential Distributions.
Standard deviation of this distribution. Standard deviation of this distribution.
- - -
#### `tf.contrib.distributions.Exponential.strict` {#Exponential.strict}
Boolean describing behavior on invalid input.
- - - - - -
#### `tf.contrib.distributions.Exponential.variance(name='variance')` {#Exponential.variance} #### `tf.contrib.distributions.Exponential.variance(name='variance')` {#Exponential.variance}
...@@ -1683,7 +1753,7 @@ dist2 = Gamma(alpha=[3.0, 4.0], beta=[2.0, 3.0]) ...@@ -1683,7 +1753,7 @@ dist2 = Gamma(alpha=[3.0, 4.0], beta=[2.0, 3.0])
``` ```
- - - - - -
#### `tf.contrib.distributions.Gamma.__init__(alpha, beta, name='Gamma')` {#Gamma.__init__} #### `tf.contrib.distributions.Gamma.__init__(alpha, beta, strict=True, name='Gamma')` {#Gamma.__init__}
Construct Gamma distributions with parameters `alpha` and `beta`. Construct Gamma distributions with parameters `alpha` and `beta`.
...@@ -1699,6 +1769,9 @@ broadcasting (e.g. `alpha + beta` is a valid operation). ...@@ -1699,6 +1769,9 @@ broadcasting (e.g. `alpha + beta` is a valid operation).
* <b>`beta`</b>: `float` or `double` tensor, the inverse scale params of the * <b>`beta`</b>: `float` or `double` tensor, the inverse scale params of the
distribution(s). distribution(s).
beta must contain only positive values. beta must contain only positive values.
* <b>`strict`</b>: Whether to assert that `a > 0, b > 0`, and that `x > 0` in the
methods `pdf(x)` and `log_pdf(x)`. If `strict` is False
and the inputs are invalid, correct behavior is not guaranteed.
* <b>`name`</b>: The name to prepend to all ops created by this distribution. * <b>`name`</b>: The name to prepend to all ops created by this distribution.
##### Raises: ##### Raises:
...@@ -1962,6 +2035,13 @@ See the doc for tf.random_gamma for further detail. ...@@ -1962,6 +2035,13 @@ See the doc for tf.random_gamma for further detail.
Standard deviation of this distribution. Standard deviation of this distribution.
- - -
#### `tf.contrib.distributions.Gamma.strict` {#Gamma.strict}
Boolean describing behavior on invalid input.
- - - - - -
#### `tf.contrib.distributions.Gamma.variance(name='variance')` {#Gamma.variance} #### `tf.contrib.distributions.Gamma.variance(name='variance')` {#Gamma.variance}
...@@ -2018,7 +2098,7 @@ dist.pdf(3.0) ...@@ -2018,7 +2098,7 @@ dist.pdf(3.0)
``` ```
- - - - - -
#### `tf.contrib.distributions.Normal.__init__(mu, sigma, name='Normal')` {#Normal.__init__} #### `tf.contrib.distributions.Normal.__init__(mu, sigma, strict=True, name='Normal')` {#Normal.__init__}
Construct Normal distributions with mean and stddev `mu` and `sigma`. Construct Normal distributions with mean and stddev `mu` and `sigma`.
...@@ -2031,6 +2111,8 @@ broadcasting (e.g. `mu + sigma` is a valid operation). ...@@ -2031,6 +2111,8 @@ broadcasting (e.g. `mu + sigma` is a valid operation).
* <b>`mu`</b>: `float` or `double` tensor, the means of the distribution(s). * <b>`mu`</b>: `float` or `double` tensor, the means of the distribution(s).
* <b>`sigma`</b>: `float` or `double` tensor, the stddevs of the distribution(s). * <b>`sigma`</b>: `float` or `double` tensor, the stddevs of the distribution(s).
sigma must contain only positive values. sigma must contain only positive values.
* <b>`strict`</b>: Whether to assert that `sigma > 0`. If `strict` is False,
correct output is not guaranteed when input is invalid.
* <b>`name`</b>: The name to give Ops created by the initializer. * <b>`name`</b>: The name to give Ops created by the initializer.
##### Raises: ##### Raises:
...@@ -2272,6 +2354,13 @@ Distribution parameter for standard deviation. ...@@ -2272,6 +2354,13 @@ Distribution parameter for standard deviation.
Standard deviation of this distribution. Standard deviation of this distribution.
- - -
#### `tf.contrib.distributions.Normal.strict` {#Normal.strict}
Boolean describing behavior on invalid input.
- - - - - -
#### `tf.contrib.distributions.Normal.variance(name='variance')` {#Normal.variance} #### `tf.contrib.distributions.Normal.variance(name='variance')` {#Normal.variance}
...@@ -2331,7 +2420,7 @@ dist.pdf(3.0) ...@@ -2331,7 +2420,7 @@ dist.pdf(3.0)
``` ```
- - - - - -
#### `tf.contrib.distributions.StudentT.__init__(df, mu, sigma, name='StudentT')` {#StudentT.__init__} #### `tf.contrib.distributions.StudentT.__init__(df, mu, sigma, strict=True, name='StudentT')` {#StudentT.__init__}
Construct Student's t distributions. Construct Student's t distributions.
...@@ -2349,6 +2438,8 @@ broadcasting (e.g. `df + mu + sigma` is a valid operation). ...@@ -2349,6 +2438,8 @@ broadcasting (e.g. `df + mu + sigma` is a valid operation).
* <b>`sigma`</b>: `float` or `double` tensor, the scaling factor for the * <b>`sigma`</b>: `float` or `double` tensor, the scaling factor for the
distribution(s). `sigma` must contain only positive values. distribution(s). `sigma` must contain only positive values.
Note that `sigma` is not the standard deviation of this distribution. Note that `sigma` is not the standard deviation of this distribution.
* <b>`strict`</b>: Whether to assert that `df > 0, sigma > 0`. If `strict` is False
and inputs are invalid, correct behavior is not guaranteed.
* <b>`name`</b>: The name to give Ops created by the initializer. * <b>`name`</b>: The name to give Ops created by the initializer.
##### Raises: ##### Raises:
...@@ -2543,6 +2634,13 @@ Scaling factors of these Student's t distribution(s). ...@@ -2543,6 +2634,13 @@ Scaling factors of these Student's t distribution(s).
- - -
#### `tf.contrib.distributions.StudentT.strict` {#StudentT.strict}
Boolean describing behavior on invalid input.
- - - - - -
#### `tf.contrib.distributions.StudentT.variance(name='variance')` {#StudentT.variance} #### `tf.contrib.distributions.StudentT.variance(name='variance')` {#StudentT.variance}
...@@ -2560,7 +2658,7 @@ Uniform distribution with `a` and `b` parameters. ...@@ -2560,7 +2658,7 @@ Uniform distribution with `a` and `b` parameters.
The PDF of this distribution is constant between [`a`, `b`], and 0 elsewhere. The PDF of this distribution is constant between [`a`, `b`], and 0 elsewhere.
- - - - - -
#### `tf.contrib.distributions.Uniform.__init__(a=0.0, b=1.0, name='Uniform')` {#Uniform.__init__} #### `tf.contrib.distributions.Uniform.__init__(a=0.0, b=1.0, strict=True, name='Uniform')` {#Uniform.__init__}
Construct Uniform distributions with `a` and `b`. Construct Uniform distributions with `a` and `b`.
...@@ -2590,12 +2688,14 @@ u1 = Uniform(3.0, [5.0, 6.0, 7.0]) # 3 distributions ...@@ -2590,12 +2688,14 @@ u1 = Uniform(3.0, [5.0, 6.0, 7.0]) # 3 distributions
* <b>`a`</b>: `float` or `double` tensor, the minimum endpoint. * <b>`a`</b>: `float` or `double` tensor, the minimum endpoint.
* <b>`b`</b>: `float` or `double` tensor, the maximum endpoint. Must be > `a`. * <b>`b`</b>: `float` or `double` tensor, the maximum endpoint. Must be > `a`.
* <b>`strict`</b>: Whether to assert that `a > b`. If `strict` is False and inputs
are invalid, correct behavior is not guaranteed.
* <b>`name`</b>: The name to prefix Ops created by this distribution class. * <b>`name`</b>: The name to prefix Ops created by this distribution class.
##### Raises: ##### Raises:
* <b>`InvalidArgumentError`</b>: if `a >= b`. * <b>`InvalidArgumentError`</b>: if `a >= b` and `strict=True`.
- - - - - -
...@@ -2785,6 +2885,13 @@ Sample `n` observations from the Uniform Distributions. ...@@ -2785,6 +2885,13 @@ Sample `n` observations from the Uniform Distributions.
- - -
#### `tf.contrib.distributions.Uniform.strict` {#Uniform.strict}
Boolean describing behavior on invalid input.
- - - - - -
#### `tf.contrib.distributions.Uniform.variance(name='variance')` {#Uniform.variance} #### `tf.contrib.distributions.Uniform.variance(name='variance')` {#Uniform.variance}
...@@ -3092,7 +3199,7 @@ dist.pmf(counts) # Shape [2] ...@@ -3092,7 +3199,7 @@ dist.pmf(counts) # Shape [2]
``` ```
- - - - - -
#### `tf.contrib.distributions.DirichletMultinomial.__init__(n, alpha, name='DirichletMultinomial', allow_arbitrary_counts=False)` {#DirichletMultinomial.__init__} #### `tf.contrib.distributions.DirichletMultinomial.__init__(n, alpha, allow_arbitrary_counts=False, strict=True, name='DirichletMultinomial')` {#DirichletMultinomial.__init__}
Initialize a batch of DirichletMultinomial distributions. Initialize a batch of DirichletMultinomial distributions.
...@@ -3106,11 +3213,12 @@ Initialize a batch of DirichletMultinomial distributions. ...@@ -3106,11 +3213,12 @@ Initialize a batch of DirichletMultinomial distributions.
* <b>`alpha`</b>: Positive `float` or `double` tensor with shape broadcastable to * <b>`alpha`</b>: Positive `float` or `double` tensor with shape broadcastable to
`[N1,..., Nm, k]` `m >= 0`. Defines this as a batch of `N1 x ... x Nm` `[N1,..., Nm, k]` `m >= 0`. Defines this as a batch of `N1 x ... x Nm`
different `k` class Dirichlet multinomial distributions. different `k` class Dirichlet multinomial distributions.
* <b>`name`</b>: The name to prefix Ops created by this distribution class.
* <b>`allow_arbitrary_counts`</b>: Boolean. This represents whether the pmf/cdf * <b>`allow_arbitrary_counts`</b>: Boolean. This represents whether the pmf/cdf
allows for the `counts` tensor to be non-integral values. allows for the `counts` tensor to be non-integral values.
The pmf/cdf are functions that can be evaluated at non-integral values, The pmf/cdf are functions that can be evaluated at non-integral values,
but are only a distribution over non-negative integers. but are only a distribution over non-negative integers.
* <b>`strict`</b>: Not used (yet).
* <b>`name`</b>: The name to prefix Ops created by this distribution class.
* <b>`Examples`</b>: * <b>`Examples`</b>:
...@@ -3339,6 +3447,13 @@ Generate `n` samples. ...@@ -3339,6 +3447,13 @@ Generate `n` samples.
Standard deviation of the distribution. Standard deviation of the distribution.
- - -
#### `tf.contrib.distributions.DirichletMultinomial.strict` {#DirichletMultinomial.strict}
Boolean describing behavior on invalid input.
- - - - - -
#### `tf.contrib.distributions.DirichletMultinomial.variance(name='variance')` {#DirichletMultinomial.variance} #### `tf.contrib.distributions.DirichletMultinomial.variance(name='variance')` {#DirichletMultinomial.variance}
......
...@@ -232,6 +232,13 @@ Standard deviation of the distribution. ...@@ -232,6 +232,13 @@ Standard deviation of the distribution.
* <b>`std`</b>: `Tensor` of the same type and shape as `p`. * <b>`std`</b>: `Tensor` of the same type and shape as `p`.
- - -
#### `tf.contrib.distributions.Bernoulli.strict` {#Bernoulli.strict}
Boolean describing behavior on invalid input.
- - - - - -
#### `tf.contrib.distributions.Bernoulli.variance(name='variance')` {#Bernoulli.variance} #### `tf.contrib.distributions.Bernoulli.variance(name='variance')` {#Bernoulli.variance}
......
...@@ -45,7 +45,7 @@ dist.pdf(3.0) ...@@ -45,7 +45,7 @@ dist.pdf(3.0)
``` ```
- - - - - -
#### `tf.contrib.distributions.StudentT.__init__(df, mu, sigma, name='StudentT')` {#StudentT.__init__} #### `tf.contrib.distributions.StudentT.__init__(df, mu, sigma, strict=True, name='StudentT')` {#StudentT.__init__}
Construct Student's t distributions. Construct Student's t distributions.
...@@ -63,6 +63,8 @@ broadcasting (e.g. `df + mu + sigma` is a valid operation). ...@@ -63,6 +63,8 @@ broadcasting (e.g. `df + mu + sigma` is a valid operation).
* <b>`sigma`</b>: `float` or `double` tensor, the scaling factor for the * <b>`sigma`</b>: `float` or `double` tensor, the scaling factor for the
distribution(s). `sigma` must contain only positive values. distribution(s). `sigma` must contain only positive values.
Note that `sigma` is not the standard deviation of this distribution. Note that `sigma` is not the standard deviation of this distribution.
* <b>`strict`</b>: Whether to assert that `df > 0, sigma > 0`. If `strict` is False
and inputs are invalid, correct behavior is not guaranteed.
* <b>`name`</b>: The name to give Ops created by the initializer. * <b>`name`</b>: The name to give Ops created by the initializer.
##### Raises: ##### Raises:
...@@ -257,6 +259,13 @@ Scaling factors of these Student's t distribution(s). ...@@ -257,6 +259,13 @@ Scaling factors of these Student's t distribution(s).
- - -
#### `tf.contrib.distributions.StudentT.strict` {#StudentT.strict}
Boolean describing behavior on invalid input.
- - - - - -
#### `tf.contrib.distributions.StudentT.variance(name='variance')` {#StudentT.variance} #### `tf.contrib.distributions.StudentT.variance(name='variance')` {#StudentT.variance}
......
...@@ -9,19 +9,20 @@ Note, the following methods of the base class aren't implemented: ...@@ -9,19 +9,20 @@ Note, the following methods of the base class aren't implemented:
* log_cdf * log_cdf
- - - - - -
#### `tf.contrib.distributions.Categorical.__init__(logits, name='Categorical', dtype=tf.int32)` {#Categorical.__init__} #### `tf.contrib.distributions.Categorical.__init__(logits, dtype=tf.int32, strict=True, name='Categorical')` {#Categorical.__init__}
Initialize Categorical distributions using class log-probabilities. Initialize Categorical distributions using class log-probabilities.
##### Args: ##### Args:
* <b>`logits`</b>: An N-D `Tensor` representing the log probabilities of a set of * <b>`logits`</b>: An N-D `Tensor`, `N >= 1`, representing the log probabilities
Categorical distributions. The first N - 1 dimensions index into a of a set of Categorical distributions. The first `N - 1` dimensions
batch of independent distributions and the last dimension indexes index into a batch of independent distributions and the last dimension
into the classes. indexes into the classes.
* <b>`name`</b>: A name for this distribution (optional).
* <b>`dtype`</b>: The type of the event samples (default: int32). * <b>`dtype`</b>: The type of the event samples (default: int32).
* <b>`strict`</b>: Unused in this distribution.
* <b>`name`</b>: A name for this distribution (optional).
- - - - - -
...@@ -188,6 +189,13 @@ Sample `n` observations from the Categorical distribution. ...@@ -188,6 +189,13 @@ Sample `n` observations from the Categorical distribution.
Standard deviation of the distribution. Standard deviation of the distribution.
- - -
#### `tf.contrib.distributions.Categorical.strict` {#Categorical.strict}
Boolean describing behavior on invalid input.
- - - - - -
#### `tf.contrib.distributions.Categorical.variance(name='variance')` {#Categorical.variance} #### `tf.contrib.distributions.Categorical.variance(name='variance')` {#Categorical.variance}
......
...@@ -2,16 +2,26 @@ The Chi2 distribution with degrees of freedom df. ...@@ -2,16 +2,26 @@ The Chi2 distribution with degrees of freedom df.
The PDF of this distribution is: The PDF of this distribution is:
```pdf(x) = (x^(df/2 - 1)e^(-x/2))/(2^(k/2)Gamma(k/2)), x > 0``` ```pdf(x) = (x^(df/2 - 1)e^(-x/2))/(2^(df/2)Gamma(df/2)), x > 0```
Note that the Chi2 distribution is a special case of the Gamma distribution, Note that the Chi2 distribution is a special case of the Gamma distribution,
with Chi2(df) = Gamma(df/2, 1/2). with Chi2(df) = Gamma(df/2, 1/2).
- - - - - -
#### `tf.contrib.distributions.Chi2.__init__(df, name='Chi2')` {#Chi2.__init__} #### `tf.contrib.distributions.Chi2.__init__(df, strict=True, name='Chi2')` {#Chi2.__init__}
Construct Chi2 distributions with parameter `df`.
##### Args:
* <b>`df`</b>: `float` or `double` tensor, the degrees of freedom of the
distribution(s). `df` must contain only positive values.
* <b>`strict`</b>: Whether to assert that `df > 0`, and that `x > 0` in the
methods `pdf(x)` and `log_pdf(x)`. If `strict` is False
and the inputs are invalid, correct behavior is not guaranteed.
* <b>`name`</b>: The name to prepend to all ops created by this distribution.
- - - - - -
...@@ -275,6 +285,13 @@ See the doc for tf.random_gamma for further detail. ...@@ -275,6 +285,13 @@ See the doc for tf.random_gamma for further detail.
Standard deviation of this distribution. Standard deviation of this distribution.
- - -
#### `tf.contrib.distributions.Chi2.strict` {#Chi2.strict}
Boolean describing behavior on invalid input.
- - - - - -
#### `tf.contrib.distributions.Chi2.variance(name='variance')` {#Chi2.variance} #### `tf.contrib.distributions.Chi2.variance(name='variance')` {#Chi2.variance}
......
...@@ -3,7 +3,7 @@ Uniform distribution with `a` and `b` parameters. ...@@ -3,7 +3,7 @@ Uniform distribution with `a` and `b` parameters.
The PDF of this distribution is constant between [`a`, `b`], and 0 elsewhere. The PDF of this distribution is constant between [`a`, `b`], and 0 elsewhere.
- - - - - -
#### `tf.contrib.distributions.Uniform.__init__(a=0.0, b=1.0, name='Uniform')` {#Uniform.__init__} #### `tf.contrib.distributions.Uniform.__init__(a=0.0, b=1.0, strict=True, name='Uniform')` {#Uniform.__init__}
Construct Uniform distributions with `a` and `b`. Construct Uniform distributions with `a` and `b`.
...@@ -33,12 +33,14 @@ u1 = Uniform(3.0, [5.0, 6.0, 7.0]) # 3 distributions ...@@ -33,12 +33,14 @@ u1 = Uniform(3.0, [5.0, 6.0, 7.0]) # 3 distributions
* <b>`a`</b>: `float` or `double` tensor, the minimum endpoint. * <b>`a`</b>: `float` or `double` tensor, the minimum endpoint.
* <b>`b`</b>: `float` or `double` tensor, the maximum endpoint. Must be > `a`. * <b>`b`</b>: `float` or `double` tensor, the maximum endpoint. Must be > `a`.
* <b>`strict`</b>: Whether to assert that `a > b`. If `strict` is False and inputs
are invalid, correct behavior is not guaranteed.
* <b>`name`</b>: The name to prefix Ops created by this distribution class. * <b>`name`</b>: The name to prefix Ops created by this distribution class.
##### Raises: ##### Raises:
* <b>`InvalidArgumentError`</b>: if `a >= b`. * <b>`InvalidArgumentError`</b>: if `a >= b` and `strict=True`.
- - - - - -
...@@ -228,6 +230,13 @@ Sample `n` observations from the Uniform Distributions. ...@@ -228,6 +230,13 @@ Sample `n` observations from the Uniform Distributions.
- - -
#### `tf.contrib.distributions.Uniform.strict` {#Uniform.strict}
Boolean describing behavior on invalid input.
- - - - - -
#### `tf.contrib.distributions.Uniform.variance(name='variance')` {#Uniform.variance} #### `tf.contrib.distributions.Uniform.variance(name='variance')` {#Uniform.variance}
......
...@@ -172,6 +172,13 @@ Generate `n` samples. ...@@ -172,6 +172,13 @@ Generate `n` samples.
Standard deviation of the distribution. Standard deviation of the distribution.
- - -
#### `tf.contrib.distributions.ContinuousDistribution.strict` {#ContinuousDistribution.strict}
Boolean describing behavior on invalid input.
- - - - - -
#### `tf.contrib.distributions.ContinuousDistribution.variance(name='variance')` {#ContinuousDistribution.variance} #### `tf.contrib.distributions.ContinuousDistribution.variance(name='variance')` {#ContinuousDistribution.variance}
......
...@@ -67,7 +67,7 @@ dist.pmf(counts) # Shape [2] ...@@ -67,7 +67,7 @@ dist.pmf(counts) # Shape [2]
``` ```
- - - - - -
#### `tf.contrib.distributions.DirichletMultinomial.__init__(n, alpha, name='DirichletMultinomial', allow_arbitrary_counts=False)` {#DirichletMultinomial.__init__} #### `tf.contrib.distributions.DirichletMultinomial.__init__(n, alpha, allow_arbitrary_counts=False, strict=True, name='DirichletMultinomial')` {#DirichletMultinomial.__init__}
Initialize a batch of DirichletMultinomial distributions. Initialize a batch of DirichletMultinomial distributions.
...@@ -81,11 +81,12 @@ Initialize a batch of DirichletMultinomial distributions. ...@@ -81,11 +81,12 @@ Initialize a batch of DirichletMultinomial distributions.
* <b>`alpha`</b>: Positive `float` or `double` tensor with shape broadcastable to * <b>`alpha`</b>: Positive `float` or `double` tensor with shape broadcastable to
`[N1,..., Nm, k]` `m >= 0`. Defines this as a batch of `N1 x ... x Nm` `[N1,..., Nm, k]` `m >= 0`. Defines this as a batch of `N1 x ... x Nm`
different `k` class Dirichlet multinomial distributions. different `k` class Dirichlet multinomial distributions.
* <b>`name`</b>: The name to prefix Ops created by this distribution class.
* <b>`allow_arbitrary_counts`</b>: Boolean. This represents whether the pmf/cdf * <b>`allow_arbitrary_counts`</b>: Boolean. This represents whether the pmf/cdf
allows for the `counts` tensor to be non-integral values. allows for the `counts` tensor to be non-integral values.
The pmf/cdf are functions that can be evaluated at non-integral values, The pmf/cdf are functions that can be evaluated at non-integral values,
but are only a distribution over non-negative integers. but are only a distribution over non-negative integers.
* <b>`strict`</b>: Not used (yet).
* <b>`name`</b>: The name to prefix Ops created by this distribution class.
* <b>`Examples`</b>: * <b>`Examples`</b>:
...@@ -314,6 +315,13 @@ Generate `n` samples. ...@@ -314,6 +315,13 @@ Generate `n` samples.
Standard deviation of the distribution. Standard deviation of the distribution.
- - -
#### `tf.contrib.distributions.DirichletMultinomial.strict` {#DirichletMultinomial.strict}
Boolean describing behavior on invalid input.
- - - - - -
#### `tf.contrib.distributions.DirichletMultinomial.variance(name='variance')` {#DirichletMultinomial.variance} #### `tf.contrib.distributions.DirichletMultinomial.variance(name='variance')` {#DirichletMultinomial.variance}
......
...@@ -8,10 +8,20 @@ Note that the Exponential distribution is a special case of the Gamma ...@@ -8,10 +8,20 @@ Note that the Exponential distribution is a special case of the Gamma
distribution, with Exponential(lam) = Gamma(1, lam). distribution, with Exponential(lam) = Gamma(1, lam).
- - - - - -
#### `tf.contrib.distributions.Exponential.__init__(lam, name='Exponential')` {#Exponential.__init__} #### `tf.contrib.distributions.Exponential.__init__(lam, strict=True, name='Exponential')` {#Exponential.__init__}
Construct Exponential distribution with parameter `lam`.
##### Args:
* <b>`lam`</b>: `float` or `double` tensor, the rate of the distribution(s).
`lam` must contain only positive values.
* <b>`strict`</b>: Whether to assert that `lam > 0`, and that `x > 0` in the
methods `pdf(x)` and `log_pdf(x)`. If `strict` is False
and the inputs are invalid, correct behavior is not guaranteed.
* <b>`name`</b>: The name to prepend to all ops created by this distribution.
- - - - - -
...@@ -272,6 +282,13 @@ Sample `n` observations from the Exponential Distributions. ...@@ -272,6 +282,13 @@ Sample `n` observations from the Exponential Distributions.
Standard deviation of this distribution. Standard deviation of this distribution.
- - -
#### `tf.contrib.distributions.Exponential.strict` {#Exponential.strict}
Boolean describing behavior on invalid input.
- - - - - -
#### `tf.contrib.distributions.Exponential.variance(name='variance')` {#Exponential.variance} #### `tf.contrib.distributions.Exponential.variance(name='variance')` {#Exponential.variance}
......
...@@ -20,7 +20,7 @@ dist2 = Gamma(alpha=[3.0, 4.0], beta=[2.0, 3.0]) ...@@ -20,7 +20,7 @@ dist2 = Gamma(alpha=[3.0, 4.0], beta=[2.0, 3.0])
``` ```
- - - - - -
#### `tf.contrib.distributions.Gamma.__init__(alpha, beta, name='Gamma')` {#Gamma.__init__} #### `tf.contrib.distributions.Gamma.__init__(alpha, beta, strict=True, name='Gamma')` {#Gamma.__init__}
Construct Gamma distributions with parameters `alpha` and `beta`. Construct Gamma distributions with parameters `alpha` and `beta`.
...@@ -36,6 +36,9 @@ broadcasting (e.g. `alpha + beta` is a valid operation). ...@@ -36,6 +36,9 @@ broadcasting (e.g. `alpha + beta` is a valid operation).
* <b>`beta`</b>: `float` or `double` tensor, the inverse scale params of the * <b>`beta`</b>: `float` or `double` tensor, the inverse scale params of the
distribution(s). distribution(s).
beta must contain only positive values. beta must contain only positive values.
* <b>`strict`</b>: Whether to assert that `a > 0, b > 0`, and that `x > 0` in the
methods `pdf(x)` and `log_pdf(x)`. If `strict` is False
and the inputs are invalid, correct behavior is not guaranteed.
* <b>`name`</b>: The name to prepend to all ops created by this distribution. * <b>`name`</b>: The name to prepend to all ops created by this distribution.
##### Raises: ##### Raises:
...@@ -299,6 +302,13 @@ See the doc for tf.random_gamma for further detail. ...@@ -299,6 +302,13 @@ See the doc for tf.random_gamma for further detail.
Standard deviation of this distribution. Standard deviation of this distribution.
- - -
#### `tf.contrib.distributions.Gamma.strict` {#Gamma.strict}
Boolean describing behavior on invalid input.
- - - - - -
#### `tf.contrib.distributions.Gamma.variance(name='variance')` {#Gamma.variance} #### `tf.contrib.distributions.Gamma.variance(name='variance')` {#Gamma.variance}
......
### `tf.pack(values, name='pack')` {#pack} ### `tf.pack(values, axis=0, name='pack')` {#pack}
Packs a list of rank-`R` tensors into one rank-`(R+1)` tensor. Packs a list of rank-`R` tensors into one rank-`(R+1)` tensor.
...@@ -14,6 +14,8 @@ This is the opposite of unpack. The numpy equivalent is ...@@ -14,6 +14,8 @@ This is the opposite of unpack. The numpy equivalent is
* <b>`values`</b>: A list of `Tensor` objects with the same shape and type. * <b>`values`</b>: A list of `Tensor` objects with the same shape and type.
* <b>`axis`</b>: An `int`. The axis to pack along. Defaults to the first dimension.
Supports negative indexes.
* <b>`name`</b>: A name for this operation (optional). * <b>`name`</b>: A name for this operation (optional).
##### Returns: ##### Returns:
...@@ -21,3 +23,8 @@ This is the opposite of unpack. The numpy equivalent is ...@@ -21,3 +23,8 @@ This is the opposite of unpack. The numpy equivalent is
* <b>`output`</b>: A packed `Tensor` with the same type as `values`. * <b>`output`</b>: A packed `Tensor` with the same type as `values`.
##### Raises:
* <b>`ValueError`</b>: If `axis` is out of the range [-(R+1), R+1).
### `tf.unpack(value, num=None, name='unpack')` {#unpack} ### `tf.unpack(value, num=None, axis=0, name='unpack')` {#unpack}
Unpacks the outer dimension of a rank-`R` tensor into rank-`(R-1)` tensors. Unpacks the given dimension of a rank-`R` tensor into rank-`(R-1)` tensors.
Unpacks `num` tensors from `value` along the first dimension. Unpacks `num` tensors from `value` along the given dimension.
If `num` is not specified (the default), it is inferred from `value`'s shape. If `num` is not specified (the default), it is inferred from `value`'s shape.
If `value.shape[0]` is not known, `ValueError` is raised. If `value.shape[axis]` is not known, `ValueError` is raised.
The ith tensor in `output` is the slice `value[i, ...]`. Each tensor in The ith tensor in `output` is the slice `value[i, ...]`. Each tensor in
`output` has shape `value.shape[1:]`. `output` has shape `value.shape[1:]`.
...@@ -17,8 +17,10 @@ This is the opposite of pack. The numpy equivalent is ...@@ -17,8 +17,10 @@ This is the opposite of pack. The numpy equivalent is
* <b>`value`</b>: A rank `R > 0` `Tensor` to be unpacked. * <b>`value`</b>: A rank `R > 0` `Tensor` to be unpacked.
* <b>`num`</b>: An `int`. The first dimension of value. Automatically inferred if * <b>`num`</b>: An `int`. The length of the dimension `axis`. Automatically inferred
`None` (the default). if `None` (the default).
* <b>`axis`</b>: An `int`. The axis to unpack along. Defaults to the first
dimension. Supports negative indexes.
* <b>`name`</b>: A name for the operation (optional). * <b>`name`</b>: A name for the operation (optional).
##### Returns: ##### Returns:
...@@ -29,4 +31,5 @@ This is the opposite of pack. The numpy equivalent is ...@@ -29,4 +31,5 @@ This is the opposite of pack. The numpy equivalent is
* <b>`ValueError`</b>: If `num` is unspecified and cannot be inferred. * <b>`ValueError`</b>: If `num` is unspecified and cannot be inferred.
* <b>`ValueError`</b>: If `axis` is out of the range [-R, R).
...@@ -237,6 +237,13 @@ Generate `n` samples. ...@@ -237,6 +237,13 @@ Generate `n` samples.
Standard deviation of the distribution. Standard deviation of the distribution.
- - -
#### `tf.contrib.distributions.BaseDistribution.strict` {#BaseDistribution.strict}
Boolean describing behavior on invalid input.
- - - - - -
#### `tf.contrib.distributions.BaseDistribution.variance(name='variance')` {#BaseDistribution.variance} #### `tf.contrib.distributions.BaseDistribution.variance(name='variance')` {#BaseDistribution.variance}
......
...@@ -158,6 +158,13 @@ Generate `n` samples. ...@@ -158,6 +158,13 @@ Generate `n` samples.
Standard deviation of the distribution. Standard deviation of the distribution.
- - -
#### `tf.contrib.distributions.DiscreteDistribution.strict` {#DiscreteDistribution.strict}
Boolean describing behavior on invalid input.
- - - - - -
#### `tf.contrib.distributions.DiscreteDistribution.variance(name='variance')` {#DiscreteDistribution.variance} #### `tf.contrib.distributions.DiscreteDistribution.variance(name='variance')` {#DiscreteDistribution.variance}
......
...@@ -42,7 +42,7 @@ dist.pdf(3.0) ...@@ -42,7 +42,7 @@ dist.pdf(3.0)
``` ```
- - - - - -
#### `tf.contrib.distributions.Normal.__init__(mu, sigma, name='Normal')` {#Normal.__init__} #### `tf.contrib.distributions.Normal.__init__(mu, sigma, strict=True, name='Normal')` {#Normal.__init__}
Construct Normal distributions with mean and stddev `mu` and `sigma`. Construct Normal distributions with mean and stddev `mu` and `sigma`.
...@@ -55,6 +55,8 @@ broadcasting (e.g. `mu + sigma` is a valid operation). ...@@ -55,6 +55,8 @@ broadcasting (e.g. `mu + sigma` is a valid operation).
* <b>`mu`</b>: `float` or `double` tensor, the means of the distribution(s). * <b>`mu`</b>: `float` or `double` tensor, the means of the distribution(s).
* <b>`sigma`</b>: `float` or `double` tensor, the stddevs of the distribution(s). * <b>`sigma`</b>: `float` or `double` tensor, the stddevs of the distribution(s).
sigma must contain only positive values. sigma must contain only positive values.
* <b>`strict`</b>: Whether to assert that `sigma > 0`. If `strict` is False,
correct output is not guaranteed when input is invalid.
* <b>`name`</b>: The name to give Ops created by the initializer. * <b>`name`</b>: The name to give Ops created by the initializer.
##### Raises: ##### Raises:
...@@ -296,6 +298,13 @@ Distribution parameter for standard deviation. ...@@ -296,6 +298,13 @@ Distribution parameter for standard deviation.
Standard deviation of this distribution. Standard deviation of this distribution.
- - -
#### `tf.contrib.distributions.Normal.strict` {#Normal.strict}
Boolean describing behavior on invalid input.
- - - - - -
#### `tf.contrib.distributions.Normal.variance(name='variance')` {#Normal.variance} #### `tf.contrib.distributions.Normal.variance(name='variance')` {#Normal.variance}
......
Markdown is supported
0% .
You are about to add 0 people to the discussion. Proceed with caution.
先完成此消息的编辑!
想要评论请 注册