"""File with Gradient SHAP algorithm explainer classes.
Based on https://github.com/pytorch/captum/blob/master/captum/attr/_core/gradient_shap.py
and https://github.com/pytorch/captum/blob/master/captum/attr/_core/layer/layer_gradient_shap.py.
"""
from abc import abstractmethod
from typing import Any, Optional, Tuple, Union
import torch
from captum._utils.typing import TargetType
from captum.attr import GradientShap, LayerGradientShap
from foxai.array_utils import validate_result
from foxai.explainer.base_explainer import Explainer
from foxai.explainer.computer_vision.model_utils import get_last_conv_model_layer
[docs]class BaseGradientSHAPCVExplainer(Explainer):
"""Base Gradient SHAP algorithm explainer."""
[docs] @abstractmethod
def create_explainer(
self,
model: torch.nn.Module,
multiply_by_inputs: bool = True,
**kwargs,
) -> Union[GradientShap, LayerGradientShap]:
"""Create explainer object.
Args:
model: The forward function of the model or any
modification of it.
multiply_by_inputs: Indicates whether to factor
model inputs' multiplier in the final attribution scores.
In the literature this is also known as local vs global
attribution. If inputs' multiplier isn't factored in
then this type of attribution method is also called local
attribution. If it is, then that type of attribution
method is called global.
More detailed can be found here:
https://arxiv.org/abs/1711.06104
In case of gradient shap, if `multiply_by_inputs`
is set to True, the sensitivity scores of scaled inputs
are being multiplied by (inputs - baselines).
Returns:
Explainer object.
"""
[docs] def calculate_features(
self,
model: torch.nn.Module,
input_data: torch.Tensor,
pred_label_idx: TargetType = None,
baselines: Union[None, int, float, torch.Tensor] = None,
n_samples: int = 5,
stdevs: Union[float, Tuple[float, ...]] = 0.0,
additional_forward_args: Any = None,
attribute_to_layer_input: bool = False,
**kwargs,
) -> torch.Tensor:
"""Generate model's attributes with Gradient SHAP algorithm explainer.
Args:
model: The forward function of the model or any
modification of it.
input_data: Input for which SHAP attribution
values are computed. If `forward_func` takes a single
tensor as input, a single input tensor should be provided.
pred_label_idx: Output indices for
which gradients are computed (for classification cases,
this is usually the target class).
If the network returns a scalar value per example,
no target index is necessary.
For general 2D outputs, targets can be either:
- a single integer or a tensor containing a single
integer, which is applied to all input examples
- a list of integers or a 1D tensor, with length matching
the number of examples in inputs (dim 0). Each integer
is applied as the target for the corresponding example.
For outputs with > 2 dimensions, targets can be either:
- A single tuple, which contains #output_dims - 1
elements. This target index is applied to all examples.
- A list of tuples with length equal to the number of
examples in inputs (dim 0), and each tuple containing
#output_dims - 1 elements. Each tuple is applied as the
target for the corresponding example.
Default: None
baselines:
Baselines define the starting point from which expectation
is computed and can be provided as:
- a single tensor, if inputs is a single tensor, with
the first dimension equal to the number of examples
in the baselines' distribution. The remaining dimensions
must match with input tensor's dimension starting from
the second dimension.
It is recommended that the number of samples in the baselines'
tensors is larger than one.
n_samples: The number of randomly generated examples
per sample in the input batch. Random examples are
generated by adding gaussian random noise to each sample.
Default: `5` if `n_samples` is not provided.
stdevs: The standard deviation
of gaussian noise with zero mean that is added to each
input in the batch. If `stdevs` is a single float value
then that same value is used for all inputs. If it is
a tuple, then it must have the same length as the inputs
tuple. In this case, each stdev value in the stdevs tuple
corresponds to the input with the same index in the inputs
tuple.
Default: 0.0
additional_forward_args: If the forward function
requires additional arguments other than the inputs for
which attributions should not be computed, this argument
can be provided. It can contain a tuple of ND tensors or
any arbitrary python type of any shape.
In case of the ND tensor the first dimension of the
tensor must correspond to the batch size. It will be
repeated for each `n_steps` for each randomly generated
input sample.
Note that the gradients are not computed with respect
to these arguments.
Default: None
attribute_to_layer_input (bool, optional): Indicates whether to
compute the attribution with respect to the layer input
or output. If `attribute_to_layer_input` is set to True
then the attributions will be computed with respect to
layer input, otherwise it will be computed with respect
to layer output.
Note that currently it is assumed that either the input
or the output of internal layer, depending on whether we
attribute to the input or output, is a single tensor.
Support for multiple tensors will be added later.
Default: False
Returns:
Attribution score computed based on GradientSHAP with respect
to each input feature. Attributions will always be
the same size as the provided inputs, with each value
providing the attribution of the corresponding input index.
If a single tensor is provided as inputs, a single tensor is
returned. If a tuple is provided for inputs, a tuple of
corresponding sized tensors is returned.
Raises:
RuntimeError: if attribution has shape (0).
"""
layer: Optional[torch.nn.Module] = kwargs.get("layer", None)
gradient_shap = self.create_explainer(model=model, layer=layer)
# defining baseline distribution of images
if baselines is None:
baselines = torch.randn(
(
2 * input_data.shape[0],
*input_data.shape[1:],
),
requires_grad=True,
device=input_data.device,
)
if isinstance(gradient_shap, LayerGradientShap):
attributions = gradient_shap.attribute(
input_data,
n_samples=n_samples,
stdevs=stdevs,
baselines=baselines,
target=pred_label_idx,
return_convergence_delta=False,
additional_forward_args=additional_forward_args,
attribute_to_layer_input=attribute_to_layer_input,
)
else:
attributions = gradient_shap.attribute(
input_data,
n_samples=n_samples,
stdevs=stdevs,
baselines=baselines,
target=pred_label_idx,
return_convergence_delta=False,
additional_forward_args=additional_forward_args,
)
validate_result(attributions=attributions)
return attributions
[docs]class GradientSHAPCVExplainer(BaseGradientSHAPCVExplainer):
"""Gradient SHAP algorithm explainer."""
[docs] def create_explainer(
self,
model: torch.nn.Module,
multiply_by_inputs: bool = True,
**kwargs,
) -> Union[GradientShap, LayerGradientShap]:
"""Create explainer object.
Args:
model: The forward function of the model or any
modification of it.
multiply_by_inputs: Indicates whether to factor
model inputs' multiplier in the final attribution scores.
In the literature this is also known as local vs global
attribution. If inputs' multiplier isn't factored in
then this type of attribution method is also called local
attribution. If it is, then that type of attribution
method is called global.
More detailed can be found here:
https://arxiv.org/abs/1711.06104
In case of gradient shap, if `multiply_by_inputs`
is set to True, the sensitivity scores of scaled inputs
are being multiplied by (inputs - baselines).
Returns:
Explainer object.
"""
return GradientShap(
forward_func=model,
multiply_by_inputs=multiply_by_inputs,
)
[docs]class LayerGradientSHAPCVExplainer(BaseGradientSHAPCVExplainer):
"""Layer Gradient SHAP algorithm explainer."""
[docs] def create_explainer(
self,
model: torch.nn.Module,
multiply_by_inputs: bool = True,
layer: Optional[torch.nn.Module] = None,
**kwargs,
) -> Union[GradientShap, LayerGradientShap]:
"""Create explainer object.
Uses parameter `layer` from `kwargs`. If not provided function will call
`get_last_conv_model_layer` function to obtain last `torch.nn.Conv2d` layer
from provided model.
Args:
model: The forward function of the model or any
modification of it.
layer: Layer for which attributions are computed.
Output size of attribute matches this layer's input or
output dimensions, depending on whether we attribute to
the inputs or outputs of the layer, corresponding to
attribution of each neuron in the input or output of
this layer.
Default: None
multiply_by_inputs: Indicates whether to factor
model inputs' multiplier in the final attribution scores.
In the literature this is also known as local vs global
attribution. If inputs' multiplier isn't factored in
then this type of attribution method is also called local
attribution. If it is, then that type of attribution
method is called global.
More detailed can be found here:
https://arxiv.org/abs/1711.06104
In case of layer gradient shap, if `multiply_by_inputs`
is set to True, the sensitivity scores for scaled inputs
are being multiplied by
layer activations for inputs - layer activations for baselines.
Returns:
Explainer object.
Raises:
ValueError: if model does not contain conv layers.
"""
if layer is None:
layer = get_last_conv_model_layer(model=model)
return LayerGradientShap(
forward_func=model,
layer=layer,
multiply_by_inputs=multiply_by_inputs,
)