MutualInformation#
- class ignite.metrics.MutualInformation(output_transform=<function Metric.<lambda>>, device=device(type='cpu'), skip_unrolling=False)[source]#
Calculates the mutual information between input $X$ and prediction $Y$.
$\begin{align*} I(X;Y) &= H(Y) - H(Y|X) = H \left( \frac{1}{N}\sum_{i=1}^N \hat{\mathbf{p}}_i \right) - \frac{1}{N}\sum_{i=1}^N H(\hat{\mathbf{p}}_i), \\ H(\mathbf{p}) &= -\sum_{c=1}^C p_c \log p_c. \end{align*}$where $\hat{\mathbf{p}}_i$ is the prediction probability vector for $i$-th input, and $H(\mathbf{p})$ is the entropy of $\mathbf{p}$.
Intuitively, this metric measures how well input data are clustered by classes in the feature space [1].
[1] https://proceedings.mlr.press/v70/hu17b.html
update
must receive output of the form(y_pred, y)
whiley
is not used in this metric.y_pred
is expected to be the unnormalized logits for each class. $(B, C)$ (classification) or $(B, C, ...)$ (e.g., image segmentation) shapes are allowed.
- Parameters
output_transform (Callable) – a callable that is used to transform the
Engine
’sprocess_function
’s output into the form expected by the metric. This can be useful if, for example, you have a multi-output model and you want to compute the metric with respect to one of the outputs. By default, metrics require the output as(y_pred, y)
or{'y_pred': y_pred, 'y': y}
.device (Union[str, device]) – specifies which device updates are accumulated on. Setting the metric’s device to be the same as your
update
arguments ensures theupdate
method is non-blocking. By default, CPU.skip_unrolling (bool) – specifies whether output should be unrolled before being fed to update method. Should be true for multi-output model, for example, if
y_pred
contains multi-ouput as(y_pred_a, y_pred_b)
Alternatively,output_transform
can be used to handle this.
Examples
To use with
Engine
andprocess_function
, simply attach the metric instance to the engine. The output of the engine’sprocess_function
needs to be in the format of(y_pred, y)
or{'y_pred': y_pred, 'y': y, ...}
. If not,output_tranform
can be added to the metric to transform the output into the form expected by the metric.For more information on how metric works with
Engine
, visit Attach Engine API.from collections import OrderedDict import torch from torch import nn, optim from ignite.engine import * from ignite.handlers import * from ignite.metrics import * from ignite.metrics.regression import * from ignite.utils import * # create default evaluator for doctests def eval_step(engine, batch): return batch default_evaluator = Engine(eval_step) # create default optimizer for doctests param_tensor = torch.zeros([1], requires_grad=True) default_optimizer = torch.optim.SGD([param_tensor], lr=0.1) # create default trainer for doctests # as handlers could be attached to the trainer, # each test must define his own trainer using `.. testsetup:` def get_default_trainer(): def train_step(engine, batch): return batch return Engine(train_step) # create default model for doctests default_model = nn.Sequential(OrderedDict([ ('base', nn.Linear(4, 2)), ('fc', nn.Linear(2, 1)) ])) manual_seed(666)
metric = MutualInformation() metric.attach(default_evaluator, 'mutual_information') y_true = torch.tensor([0, 1, 2]) # not considered in the MutualInformation metric. y_pred = torch.tensor([ [ 0.0000, 0.6931, 1.0986], [ 1.3863, 1.6094, 1.6094], [ 0.0000, -2.3026, -2.3026] ]) state = default_evaluator.run([[y_pred, y_true]]) print(state.metrics['mutual_information'])
0.18599730730056763
Changed in version 0.5.1:
skip_unrolling
argument is added.Methods
Computes the metric based on its accumulated state.
Resets the metric to its initial state.
- compute()[source]#
Computes the metric based on its accumulated state.
By default, this is called at the end of each epoch.
- Returns
- the actual quantity of interest. However, if a
Mapping
is returned, it will be (shallow) flattened into engine.state.metrics whencompleted()
is called. - Return type
Any
- Raises
NotComputableError – raised when the metric cannot be computed.