How mostafa-abbas/propsens works
propsens is an R package for assessing how propensity-model choice affects a matched study design, without using outcomes. The analyst supplies study data and matching settings, or an existing MatchIt object; the documented workflow refits models, rematches, applies a quality gate, and reports sensitivity diagnostics. The graph retains the README’s documented diagnostic and seed-sensitivity capabilities. Source excerpts were not provided, so relationships are limited to those stated in the README; some internal subsystem paths point to their enclosing R directory where implementation details are unavailable.
Main entry points: propensity_sensitivity.R Sensitivity run · seed_sensitivity.R Seed sensitivity · simulate_example_data.R Example data generator · fit_propensity_models.R Propensity models
Components 16
Study inputs
- Study data
- MatchIt object
Model and matching
- Sensitivity run
R/propensity_sensitivity.R - Propensity models
R/fit_propensity_models.R - Rematching
R/match_each_model.R
Sensitivity assessment
- Quality gate
R/propensity_sensitivity.R - Rank disagreement
R/rank_disagreement.R - Pair instability
R/pair_recurrence.R - Bootstrap rank instability
R/bootstrap_rank_instability.R - Balance spread
R/balance_spread.R - Control spread
R/control_spread.R
Results and tools
- Seed sensitivity
R/seed_sensitivity.R - Diagnostic results
R/propensity_sensitivity.R - Sensitivity plots
R/plot_propensity_sensitivity.R - Example data generator
R/simulate_example_data.R
Other components
- Study analyst
Connections 16
- Study analyst → Sensitivity run: provides design
README.md - Study data → Sensitivity run: supplies data
README.md - MatchIt object → Sensitivity run: supplies design
README.md - Sensitivity run → Propensity models: refits models
README.md - Sensitivity run → Rematching: repeats matching
README.md - Sensitivity run → Quality gate: applies gate
README.md - Sensitivity run → Rank disagreement: reports ranks
README.md - Sensitivity run → Pair instability: reports pairs
README.md - Sensitivity run → Bootstrap rank instability: reports bootstrap
README.md - Sensitivity run → Balance spread: reports balance
README.md - Sensitivity run → Control spread: reports spread
README.md - Sensitivity run → Diagnostic results: returns results
README.md - Seed sensitivity → Sensitivity run: reruns seeds
README.md - Study analyst → Seed sensitivity: requests comparison
README.md - Diagnostic results → Sensitivity plots: can be plotted
- Example data generator → Study data: generates example
README.md