Multivariate Support and Performance Improvements in imputeTestbench

This project focuses on extending the imputeTestbench R package to better support modern time series data. The current package works well for univariate data, but many real-world datasets are multivariate, where multiple variables evolve together over time and may have structured missing values. The main goal is to add support for multivariate time series while keeping the package simple, stable, and fully compatible with CRAN. The project will introduce methods to handle multivariate inputs, simulate realistic missing data patterns such as block missingness, and improve benchmarking so that imputation methods can be evaluated both per variable and in aggregated form. In addition, the project will improve performance through efficient data handling and optional parallel processing using CRAN-safe approaches. The internal structure of the package will also be refined to improve readability, maintainability, and future extensibility. The expected deliverables include multivariate input support, structured missing data generation, improved benchmarking functions, optional parallel execution, and updated testing and documentation. These improvements will make imputeTestbench more useful for real-world applications such as environmental data, energy systems, finance, and sensor-based monitoring.