bsvarPost is a post-estimation companion package for bsvars and bsvarSIGNs.

It focuses on what happens after estimation:

• cumulative dynamic multipliers
• tidy extraction of posterior objects
• model comparison
• representative-model summaries
• restriction, magnitude, and hypothesis audits
• diagnostics for sign-restricted models
• historical-decomposition event workflows
• publication-oriented plotting and reporting

Start Here

• New to the package: Getting Started
• Looking for the broader workflow surface: Post-Estimation Workflows

Core Workflow

1. Estimate a model with bsvars or bsvarSIGNs.
2. Extract responses or summaries with tidy_*() or cdm().
3. Compare specifications with compare_*().
4. Audit hypotheses and restrictions.
5. Move results into publication-oriented tables and plots.

Main Feature Families

Responses and extraction

• cdm()
• tidy_irf(), tidy_cdm(), tidy_fevd(), tidy_hd(), tidy_forecast()

Comparison

• compare_irf(), compare_cdm(), compare_fevd(), compare_forecast()
• comparison helpers for response-shape summaries and diagnostics

Representative summaries and audits

• median_target_*()
• most_likely_admissible_*()
• hypothesis_*()
• restriction_audit()
• magnitude_audit()

Historical decomposition workflows

• tidy_hd_event()
• shock_ranking()
• compare_hd_event()

Publication outputs

• publish_bsvar_plot()
• report_bundle()
• report_table()
• as_kable(), as_gt(), as_flextable()

Notes

• bsvarPost does not estimate models itself.
• Magnitude restrictions are handled as post-estimation audits, not estimation-time admissibility filters.
• For sign-restricted models, most_likely_admissible_*() is available only for PosteriorBSVARSIGN.