Merge pull request #848 from milaogou/helix-pr

feat: add HELIX model for time series imputation

Author: WenjieDu

.github/workflows/testing_daily.yml

@@ -27,7 +27,7 @@ jobs:
 
 
             - name: Set up Conda
-              uses: conda-incubator/setup-miniconda@v3
+              uses: conda-incubator/setup-miniconda@v4
               with:
                   miniconda-version: "latest"
                   activate-environment: pypots

pypots/imputation/__init__.py

@@ -34,6 +34,7 @@
 from .micn import MICN
 from .tide import TiDE
 from .grud import GRUD
+from .helix import HELIX
 from .stemgnn import StemGNN
 from .imputeformer import ImputeFormer
 from .timemixer import TimeMixer
@@ -85,6 +86,7 @@
     "MICN",
     "TiDE",
     "GRUD",
+    "HELIX",
     "StemGNN",
     "ImputeFormer",
     "TimeMixer",
@@ -105,4 +107,4 @@
     "Mean",
     "Median",
     "Lerp",
-]
+]

pypots/imputation/helix/__init__.py

@@ -0,0 +1,25 @@
+"""
+The package of the partially-observed time-series imputation model HELIX.
+
+Refer to the paper
+`MiBah Cat.
+HELIX: Hybrid Encoding with Learnable Identity and Cross-dimensional Synthesis 
+for Time Series Imputation.
+<paper_url_placeholder>`_
+
+Notes
+-----
+HELIX employs rotary positional encoding for temporal dimension and learnable 
+identity embeddings for feature dimension, combined with parallel and serial 
+cross-dimensional attention mechanism.
+
+"""
+
+# Created by MiBah Cat <milaogou@gmail.com>
+# License: BSD-3-Clause
+
+from .model import HELIX
+
+__all__ = [
+    "HELIX",
+]