Merge pull request #389 from WenjieDu/(feat)add_pyraformer

Implement Pyraformer as an imputation model

pypots/imputation/__init__.py

@@ -14,6 +14,7 @@
 from .transformer import Transformer
 from .itransformer import iTransformer
 from .nonstationary_transformer import NonstationaryTransformer
+from .pyraformer import Pyraformer
 from .timesnet import TimesNet
 from .etsformer import ETSformer
 from .fedformer import FEDformer
@@ -47,6 +48,7 @@
     "Informer",
     "Autoformer",
     "NonstationaryTransformer",
+    "Pyraformer",
     "BRITS",
     "MRNN",
     "GPVAE",

pypots/imputation/nonstationary_transformer/model.py

@@ -24,7 +24,7 @@
 
 class NonstationaryTransformer(BaseNNImputer):
     """The PyTorch implementation of the Nonstationary-Transformer model.
-    NonstationaryTransformer is originally proposed by Wu et al. in :cite:`liu2022nonstationary`.
+    NonstationaryTransformer is originally proposed by Liu et al. in :cite:`liu2022nonstationary`.
 
     Parameters
     ----------

pypots/imputation/pyraformer/__init__.py

@@ -0,0 +1,24 @@
+"""
+The package of the partially-observed time-series imputation model Pyraformer.
+
+Refer to the paper
+`Shizhan Liu, Hang Yu, Cong Liao, Jianguo Li, Weiyao Lin, Alex X. Liu, and Schahram Dustdar.
+"Pyraformer: Low-Complexity Pyramidal Attention for Long-Range Time Series Modeling and Forecasting".
+International Conference on Learning Representations. 2022.
+<https://openreview.net/pdf?id=0EXmFzUn5I>`_
+
+Notes
+-----
+This implementation is inspired by the official one https://github.com/ant-research/Pyraformer
+
+"""
+
+# Created by Wenjie Du <[email protected]>
+# License: BSD-3-Clause
+
+
+from .model import Pyraformer
+
+__all__ = [
+    "Pyraformer",
+]

pypots/imputation/pyraformer/core.py

@@ -0,0 +1,86 @@
+"""
+The core wrapper assembles the submodules of Pyraformer imputation model
+and takes over the forward progress of the algorithm.
+"""
+
+# Created by Wenjie Du <[email protected]>
+# License: BSD-3-Clause
+
+import torch.nn as nn
+
+from ...nn.modules.pyraformer import PyraformerEncoder
+from ...nn.modules.saits import SaitsLoss, SaitsEmbedding
+
+
+class _Pyraformer(nn.Module):
+    def __init__(
+        self,
+        n_steps: int,
+        n_features: int,
+        n_layers: int,
+        d_model: int,
+        n_heads: int,
+        d_ffn: int,
+        dropout: float,
+        attn_dropout: float,
+        window_size: list,
+        inner_size: int,
+        ORT_weight: float = 1,
+        MIT_weight: float = 1,
+    ):
+        super().__init__()
+
+        self.saits_embedding = SaitsEmbedding(
+            n_features * 2,
+            d_model,
+            with_pos=False,
+            dropout=dropout,
+        )
+        self.encoder = PyraformerEncoder(
+            n_steps,
+            n_layers,
+            d_model,
+            n_heads,
+            d_ffn,
+            dropout,
+            attn_dropout,
+            window_size,
+            inner_size,
+        )
+
+        # for the imputation task, the output dim is the same as input dim
+        self.output_projection = nn.Linear((len(window_size) + 1) * d_model, n_features)
+        self.saits_loss_func = SaitsLoss(ORT_weight, MIT_weight)
+
+    def forward(self, inputs: dict, training: bool = True) -> dict:
+        X, missing_mask = inputs["X"], inputs["missing_mask"]
+
+        # WDU: the original Pyraformer paper isn't proposed for imputation task. Hence the model doesn't take
+        # the missing mask into account, which means, in the process, the model doesn't know which part of
+        # the input data is missing, and this may hurt the model's imputation performance. Therefore, I apply the
+        # SAITS embedding method to project the concatenation of features and masks into a hidden space, as well as
+        # the output layers to project back from the hidden space to the original space.
+        enc_out = self.saits_embedding(X, missing_mask)
+
+        # Pyraformer encoder processing
+        enc_out, attns = self.encoder(enc_out)
+        # project back the original data space
+        reconstruction = self.output_projection(enc_out)
+
+        imputed_data = missing_mask * X + (1 - missing_mask) * reconstruction
+        results = {
+            "imputed_data": imputed_data,
+        }
+
+        # if in training mode, return results with losses
+        if training:
+            X_ori, indicating_mask = inputs["X_ori"], inputs["indicating_mask"]
+            loss, ORT_loss, MIT_loss = self.saits_loss_func(
+                reconstruction, X_ori, missing_mask, indicating_mask
+            )
+            results["ORT_loss"] = ORT_loss
+            results["MIT_loss"] = MIT_loss
+            # `loss` is always the item for backward propagating to update the model
+            results["loss"] = loss
+
+        return results

pypots/imputation/pyraformer/data.py

@@ -0,0 +1,24 @@
+"""
+Dataset class for Pyraformer.
+"""
+
+# Created by Wenjie Du <[email protected]>
+# License: BSD-3-Clause
+
+from typing import Union
+
+from ..saits.data import DatasetForSAITS
+
+
+class DatasetForPyraformer(DatasetForSAITS):
+    """Actually Pyraformer uses the same data strategy as SAITS, needs MIT for training."""
+
+    def __init__(
+        self,
+        data: Union[dict, str],
+        return_X_ori: bool,
+        return_y: bool,
+        file_type: str = "hdf5",
+        rate: float = 0.2,
+    ):
+        super().__init__(data, return_X_ori, return_y, file_type, rate)