pyFAST: A Modular PyTorch Framework for Time Series Modeling with Multi-source and Sparse Data

2025-08-26

AI Summary

Existing Python time-series frameworks lack native support for irregular, multi-source, and sparse temporal data, limiting modeling flexibility and scalability. To address this, we propose a modular time-series modeling framework that decouples data preprocessing from model computation, enabling irregular-sampling data loading, dynamic normalization, mask-driven modeling, and external covariate integration. We introduce an alignment-agnostic sparse data fusion mechanism supporting efficient sequence- or block-level padding, alongside dedicated sparsity-aware metrics and loss functions. Built on PyTorch, the framework unifies linear models, CNNs, RNNs, Transformers, GNNs, and LLM-inspired architectures, and incorporates batch-stream co-training and hardware-aware optimization strategies. Experiments demonstrate superior scalability and computational efficiency on multi-source heterogeneous and sparse time-series forecasting and imputation tasks.

Technology Category

• Machine Learning: Time-Series/Data Streams
• Computer Vision: Multi-modal Vision
• Data Mining & Knowledge Management: Mining of Spatial, Temporal or Spatio-Temporal Data

Application Category

• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• User Modeling, Personalization and Recommendation: Federated recommendation systems and personalization
• Systems and Infrastructure for Web, Mobile and WoT: Data management and stream processing for Web, mobile and wireless applications

Abstract

Modern time series analysis demands frameworks that are flexible, efficient, and extensible. However, many existing Python libraries exhibit limitations in modularity and in their native support for irregular, multi-source, or sparse data. We introduce pyFAST, a research-oriented PyTorch framework that explicitly decouples data processing from model computation, fostering a cleaner separation of concerns and facilitating rapid experimentation. Its data engine is engineered for complex scenarios, supporting multi-source loading, protein sequence handling, efficient sequence- and patch-level padding, dynamic normalization, and mask-based modeling for both imputation and forecasting. pyFAST integrates LLM-inspired architectures for the alignment-free fusion of sparse data sources and offers native sparse metrics, specialized loss functions, and flexible exogenous data fusion. Training utilities include batch-based streaming aggregation for evaluation and device synergy to maximize computational efficiency. A comprehensive suite of classical and deep learning models (Linears, CNNs, RNNs, Transformers, and GNNs) is provided within a modular architecture that encourages extension. Released under the MIT license at GitHub, pyFAST provides a compact yet powerful platform for advancing time series research and applications.

Problem

• Handling irregular multi-source sparse time series data
• Decoupling data processing from model computation
• Supporting alignment-free fusion of sparse data sources

Innovation

• Modular PyTorch framework for multi-source sparse data
• Decouples data processing from model computation
• Integrates LLM-inspired architectures for alignment-free fusion