Feature/cycle pooling

[eswarchandravidyasagar]

This pull request introduces a new "Cycle Pooling" analysis mode to the application, allowing users to combine two or more survey cycles into a single average-period estimate. The implementation includes UI changes, workflow logic, and backend analysis support for pooling, with clear user guidance and compatibility checks. The main logic is refactored to support both the new pooling mode and the existing multi-cycle and single-cycle workflows.

Major features and changes:

New Cycle Pooling Workflow

• Added "Cycle Pooling" as an analysis mode, which combines selected cycles into a pooled estimate with scaled weights and independent-cycle variance estimation. This includes new UI guidance, validation for minimum cycles, and informative warnings/errors for users. [1] [2] [3] [4]
• Implemented compatibility checks to ensure only harmonized variables with equivalent meanings and categories across all selected cycles are included in pooling, with user feedback and review of excluded variables.

Backend Analysis Support

• Integrated new pooling-specific data preparation and analysis functions: filter_poolable_variables, prepare_pooled_variable, and run_cycle_pooled_analysis. These handle harmonization, weight scaling, and variance estimation for pooled data. [1] [2] [3]
• Refactored batch and single-variable analysis logic to support pooled analysis, including correct labeling and results display. [1] [2] [3]

UI and Workflow Refactoring

• Unified the handling of multi-cycle and pooling modes with shared logic and variables, reducing code duplication and improving maintainability. [1] [2] [3] [4] [5] [6]
• Enhanced the results dashboard to clearly distinguish between pooled and multi-cycle results, with updated headers, badges, and explanatory text.

Documentation Updates

• Updated the README.md to document the new Cycle Pooling workflow, its statistical methodology, and user responsibilities for ensuring substantive comparability across cycles. [1] [2]