NEWS

---

FastJM 1.7.2 (2026-10-19)

• Add coef() and logLik() to extract the estimated coefficients and log likelihood values.
• Fix small bugs in summary().
• Fix coding issues in the example for running JMMLSM().
• Remove survfitjmcs(), survfitJMMLSM(), and survfitmvjmcs(), and add survfitJM() as a unified interface function to make dynamic predictions for new subjects.

FastJM 1.7.1 (2026-08-21)

• Fix small bugs in an estimation procedure.

FastJM 1.7.0 (2026-07-21)

• Unified prediction and model evaluation across all supported joint models through the new functions survfitJM(), DynPredAcc(), and Concordance().

FastJM 1.6.0 (2026-03-28)

• The package now supports modeling a single longitudinal outcome with heterogeneous within-subject variability via functionality adapted from the 'JMH' package.
• Add DynPredAccjmcs() to calculate all available evaluation metrics for fitted joint models.
• Remove AUCjmcs(), MAEQjmcs(), PEjmcs(), summary.AUCjmcs(), summary.MAEQjmcs(), and summary.PEjmcs().

FastJM 1.5.3 (2025-11-08)

• Fix small bug in survfitmvjmcs().

FastJM 1.5.2 (2025-10-18)

• Fix small bug in mvjmcs() to support ties in obtaining the estimator of the baseline hazards.

FastJM 1.5.1 (2025-07-23)

• Fix small bugs and add the estimates and 95% confidence interval for the error variances of multivariate joint modeling.

FastJM 1.5.0 (2025-06-22)

• Add a new function mvjmcs() to support multivariate joint modeling.

FastJM 1.4.2 (2024-02-29)

• Fix small bugs within C functions.

FastJM 1.4.1 (2024-01-09)

• Fix small bugs within jmcs().

FastJM 1.4.0 (2023-10-10)

• Add the function AUCjmcs() area under the ROC curve (AUC) to assess the prediction performance of joint models.

FastJM 1.3.0

• Correct the implementation of dynamic prediction in surviftjmcs().
• Remove plot.surviftjmcs() due to theoretical problem.
• Provide two metrics of prediction accuracy of joint model by adding PEjmcs() and MAEQjmcs().

FastJM 1.2.0 (2022-08-06)

• Correct the implementation of dynamic prediction in surviftjmcs() for the competing risk and add summary() for providing parameter estimates and SE for both sub-models.

FastJM 1.1.3 (2022-06-01)

• Correct syntax error on jmcs().

FastJM 1.1.2 (2022-02-16)

• Correct testthat error.

FastJM 1.1.1 (2022-02-14)

• Correct testthat error.

FastJM 1.1.0 (2022-02-08)

• Provide support for handling categorical variables in both sub-models.

• Provide the anova() function to compare two fitted joint models.

• Add the simulate argument in the survfitjmcs() function to obtain the conditional probabilities using the Gauss-Hermite quadrature rule for numerical integration.

• Adjust the label position of y axis for clarity purposes when include.y = TRUE in the plot.survfitjmcs() function.

FastJM 1.0.1 (2022-01-11)

• Removed unused variables in C functions.

FastJM 1.0.0 (2022-01-06)

• First version of software with subsequent minor patches. No NEWS file was maintained.