Changes in version 0.4.1 (2026-01-24)                  

  - Updated the standardized-effect-size vignette author list.

                 Changes in version 0.4.0 (2026-01-24)                  

  - Added standardized proximal effect size estimation for continuous
    proximal outcomes:
      - New function calculate_mrt_effect_size() with optional LOESS
        smoothing and participant-level bootstrap CIs.
      - New S3 methods summary.mrt_effect_size() and
        plot.mrt_effect_size() for concise summaries and plotting with
        CIs.
      - New example dataset data_example_for_standardized_effect to
        illustrate usage.
      - New vignette: "Standardized Proximal Effect Size in
        MRTAnalysis".
      - Updated documentation and README examples to use the new
        summary/plot methods.

                 Changes in version 0.3.1 (2025-11-05)                  

  - Fixed a bug in wcls() where input data with unordered ID may cause
    dimension errors in matrix operations. Specifically, the split()
    function was sorting IDs alphabetically while cluster sizes and
    working covariance matrices used the order IDs appeared in the data.
    Now uses factor() with explicit levels to preserve ID ordering
    across all internal functions (wcls_bread(), leverage(),
    wcls_estfun(), wcls_meat(), working.covariance()).

                 Changes in version 0.3.0 (2025-09-09)                  

  - Added new functionality for mediated causal excursion effects in
    MRTs:
      - Added mcee() function: streamlined workflow for estimating
        natural direct excursion effect (NDEE) and natural indirect
        excursion effect (NIEE) in micro-randomized trials (MRTs) with
        distal outcomes.
      - Added two advanced wrappers:
      - mcee_general(): flexible configuration of nuisance models (p, q,
        eta, mu, nu) with support for multiple learners (glm, gam, lm,
        rf, ranger, sl).
      - mcee_userfit_nuisance(): allows users to inject externally
        fitted nuisance predictions.
      - Included config helper functions (mcee_config_glm(),
        mcee_config_gam(), mcee_config_ranger(), etc.) and
        mcee_config_maker() for building nuisance specifications to pass
        into mcee_general().
      - New dataset data_time_varying_mediator_distal_outcome included
        to illustrate usage.
      - Added vignette "Time-Varying Causal Excursion Effect Mediation
        in MRT: Continuous Distal Outcomes" with detailed examples and
        best practices.

                 Changes in version 0.2.0 (2025-08-29)                  

  - Added new functionality for distal outcomes in MRTs:
      - Implemented dcee() for estimating distal causal excursion
        effects.
      - Supports flexible nuisance regression learners (lm, gam, rf,
        ranger, SuperLearner) with optional cross-fitting.
      - Provides small-sample t inference via summary.dcee_fit(),
        consistent with wcls() and emee().
      - New synthetic dataset data_distal_continuous for examples and
        testing.
      - Added vignette: Exploratory Analysis for MRT: Distal Outcomes.
  - Minor bug fixes and improvements to wcls() and emee() documentation.

                 Changes in version 0.1.2 (2023-07-01)                  

  - Fixed a bug in wcls when the randomization probability is
    time-varying.
  - Now all variable inputs need to be in quotation marks; for example,
    from now on one should specify id = "userid" instead of id = userid.
    This is to allow dynamically specified column names.

                 Changes in version 0.1.1 (2023-03-10)                  

  - Updated vignette to improve clarify.

                 Changes in version 0.1.0 (2023-03-06)                  

  - Initial release.