Changes in version 1.0.1.9001                      

                 Changes in version 1.0.1 (2025-07-28)                  

  - opsr_simulate() now takes seed argument to enable reproducibility.
  - plot.opsr() and opsr_te() define type arg list to behave similarly
    to predict.opsr().

                 Changes in version 1.0.0 (2025-07-07)                  

  - print.summary.opsr() now shows GOF indicators.
  - Added timeuse_data from the TimeUse+ study.
  - extract.opsr(): R2 for each regime in texreg functions (and don't
    repeat GOFs if beside = TRUE).
  - Nicer summary printing and setting fixed values in null model to NA.
  - Added .loglik arg to opsr() to return log-likelihood vector.
  - Detects and warns on singularity issues.
  - Resolved memory leak.
  - Corrected outcome-specific R2 computations.
  - Added type = "Xb" and type = "correction" in predict.opsr().
  - Works for only 2 treatment regimes (i.e., the Tobit-5 model).
  - Smoothing constant (used in continuity correction) can be passed to
    predict.opsr() via delta arg.
  - Added opsr_te() for treatment effect computations.
  - Added plot() method for treatment effect visualization.
  - Improved predict()'s argument argument checking and error messages
    as well as the man pages.
  - Passing the null model as sole argument to anova() will raise an
    error.
  - Added vignette corresponding to the JSS submission.

                 Changes in version 0.1.2 (2024-11-01)                  

  - Patch with minor updates on documentation.

                        Changes in version 0.1.1                        

  - Patch with missing links.

                        Changes in version 0.1.0                        

  - Initial CRAN submission.