Source code for pypots.anomaly_detection.informer.model
"""
The implementation of Informer for the partially-observed time-series anomaly detection task.
"""
# Created by Yiyuan Yang <yyy1997sjz@gmail.com>
# License: BSD-3-Clause
from typing import Union, Optional
import torch
from torch.utils.data import DataLoader
from ..base import BaseNNDetector
from ...data.checking import key_in_data_set
from ...imputation.informer.core import _Informer
from ...imputation.saits.data import DatasetForSAITS
from ...nn.modules.loss import Criterion, MAE, MSE
from ...optim.adam import Adam
from ...optim.base import Optimizer
[docs]
class Informer(BaseNNDetector):
"""The PyTorch implementation of the Informer model :cite:zhou2021informer for the anomaly detection task.
Parameters
----------
n_steps : int
The number of time steps in the time-series data sample.
n_features : int
The number of features in the time-series data sample.
anomaly_rate : float
The estimated anomaly rate in the dataset, within the range (0, 1). Used for thresholding.
n_layers : int
The number of layers in the Informer model.
d_model : int
The dimensionality of input and output feature vectors.
n_heads : int
The number of attention heads in each Transformer block.
d_ffn : int
The dimensionality of the feed-forward network hidden layer.
factor : int
The factor for controlling sparsity in the ProbSparse attention mechanism.
dropout : float, optional
Dropout rate applied throughout the model. Default is 0.
ORT_weight : float, optional
The weight for the ORT loss term during training. Default is 1.
MIT_weight : float, optional
The weight for the MIT loss term during training. Default is 1.
batch_size : int, optional
The number of samples per batch during training and evaluation. Default is 32.
epochs : int, optional
The total number of training epochs. Default is 100.
patience : int, optional
Number of epochs to wait for early stopping if no improvement. If None, early stopping is disabled.
training_loss : Criterion or type, optional
Loss function used during training. Defaults to MAE.
validation_metric : Criterion or type, optional
Metric function used for validation during training. Defaults to MSE.
optimizer : Optimizer or type, optional
The optimizer instance or optimizer class used for model training. Defaults to custom Adam optimizer.
num_workers : int, optional
The number of subprocesses to use for data loading. Default is 0.
device : str, torch.device, or list, optional
The device(s) on which the model will be trained or evaluated. Supports CPU, CUDA, and multi-GPU.
saving_path : str, optional
The path where model checkpoints and training logs are saved. No saving if None.
model_saving_strategy : str or None, optional
Strategy to save models during training: one of {None, "best", "better", "all"}.
verbose : bool, optional
Whether to print detailed logs during the training process. Default is True.
"""
def __init__(
self,
n_steps: int,
n_features: int,
anomaly_rate: float,
n_layers: int,
d_model: int,
n_heads: int,
d_ffn: int,
factor: int,
dropout: float = 0,
ORT_weight: float = 1,
MIT_weight: float = 1,
batch_size: int = 32,
epochs: int = 100,
patience: Optional[int] = None,
training_loss: Union[Criterion, type] = MAE,
validation_metric: Union[Criterion, type] = MSE,
optimizer: Union[Optimizer, type] = Adam,
num_workers: int = 0,
device: Optional[Union[str, torch.device, list]] = None,
saving_path: str = None,
model_saving_strategy: Optional[str] = "best",
verbose: bool = True,
):
# Call base detector initialization
super().__init__(
anomaly_rate=anomaly_rate,
training_loss=training_loss,
validation_metric=validation_metric,
batch_size=batch_size,
epochs=epochs,
patience=patience,
num_workers=num_workers,
device=device,
saving_path=saving_path,
model_saving_strategy=model_saving_strategy,
verbose=verbose,
)
# Save model hyperparameters
self.n_steps = n_steps
self.n_features = n_features
self.n_layers = n_layers
self.d_model = d_model
self.n_heads = n_heads
self.d_ffn = d_ffn
self.factor = factor
self.dropout = dropout
self.ORT_weight = ORT_weight
self.MIT_weight = MIT_weight
# Instantiate the Informer model
self.model = _Informer(
n_steps=self.n_steps,
n_features=self.n_features,
n_layers=self.n_layers,
d_model=self.d_model,
n_heads=self.n_heads,
d_ffn=self.d_ffn,
factor=self.factor,
dropout=self.dropout,
distil=False, # distilling disabled
ORT_weight=self.ORT_weight,
MIT_weight=self.MIT_weight,
training_loss=self.training_loss,
validation_metric=self.validation_metric,
)
# Move model to target device (GPU/CPU)
self._send_model_to_given_device()
# Print model size
self._print_model_size()
# Initialize optimizer
if isinstance(optimizer, Optimizer):
self.optimizer = optimizer
else:
self.optimizer = optimizer()
assert isinstance(self.optimizer, Optimizer)
self.optimizer.init_optimizer(self.model.parameters())
def _assemble_input_for_training(self, data: list) -> dict:
"""
Assemble input batch dictionary for model training.
Parameters
----------
data : list
A list of batched data from the DataLoader.
Returns
-------
dict
Formatted input dictionary with keys 'X', 'missing_mask', 'X_ori', and 'indicating_mask'.
"""
# Move data tensors to the correct device
indices, X, missing_mask, X_ori, indicating_mask = self._send_data_to_given_device(data)
return {
"X": X,
"missing_mask": missing_mask,
"X_ori": X_ori,
"indicating_mask": indicating_mask,
}
def _assemble_input_for_validating(self, data: list) -> dict:
"""
Assemble input batch dictionary for model validation.
Parameters
----------
data : list
A list of batched validation data.
Returns
-------
dict
Same format as training input.
"""
return self._assemble_input_for_training(data)
def _assemble_input_for_testing(self, data: list) -> dict:
"""
Assemble input batch dictionary for model testing.
Parameters
----------
data : list
A list of batched testing data.
Returns
-------
dict
Input dictionary containing 'X' and 'missing_mask' only.
"""
indices, X, missing_mask = self._send_data_to_given_device(data)
return {
"X": X,
"missing_mask": missing_mask,
}
[docs]
def fit(
self,
train_set: Union[dict, str],
val_set: Optional[Union[dict, str]] = None,
file_type: str = "hdf5",
) -> None:
"""
Train the Informer model on the provided datasets.
Parameters
----------
train_set : dict or str
The training dataset or file path.
val_set : dict or str, optional
The validation dataset or file path. Must contain 'X_ori' if provided.
file_type : str, optional
File type if data is loaded from disk. Default is "hdf5".
"""
# Save reference to training set
self.train_set = train_set
# Prepare training dataset and DataLoader
train_dataset = DatasetForSAITS(train_set, return_X_ori=False, return_y=False, file_type=file_type)
train_dataloader = DataLoader(
train_dataset,
batch_size=self.batch_size,
shuffle=True, # Shuffle during training
num_workers=self.num_workers,
)
# Prepare validation dataset and DataLoader if provided
val_dataloader = None
if val_set is not None:
if not key_in_data_set("X_ori", val_set):
raise ValueError("val_set must contain 'X_ori' for validation.")
val_dataset = DatasetForSAITS(val_set, return_X_ori=True, return_y=False, file_type=file_type)
val_dataloader = DataLoader(
val_dataset,
batch_size=self.batch_size,
shuffle=False, # No shuffle during validation
num_workers=self.num_workers,
)
# Train the model
self._train_model(train_dataloader, val_dataloader)
# Load best model state
self.model.load_state_dict(self.best_model_dict)
# Save best model if configured
self._auto_save_model_if_necessary(confirm_saving=self.model_saving_strategy == "best")