Test with Fedora 39 and Quarto 1.4.489

[XiangyunHuang]

• R 3.4.2 and others
• Matrix 1.6.2 及其下游 R 包兼容性 INLA / glmmTMB

[XiangyunHuang]

Quitting from lines 615-619 [unnamed-chunk-44] (hierarchical-normal-models.qmd)
Error in `fitTMB()`:
! negative log-likelihood is NaN at starting parameter values
Backtrace:
 1. glmmTMB::glmmTMB(...)
 2. glmmTMB::fitTMB(TMBStruc)
Execution halted

Error: Process completed with exit code 1.

在 Matrix 更新后，TMB 和 glmmTMB 最好从源码重新安装或等二者更新后从 CRAN 安装。特别是 Matrix 包发布重大版本更新后，TMB 和 glmmTMB 必须更新，否则，结果会出问题。

[XiangyunHuang]

Error in validObject(.Object) : 
  invalid class “dsparseModelMatrix” object: superclass "xMatrix" not defined in the environment of the object's class

 *** inla.core.safe:  inla.program has crashed: rerun to get better initial values. try=1/1 
Error in validObject(.Object) : 
  invalid class “dsparseModelMatrix” object: superclass "xMatrix" not defined in the environment of the object's class
Error in inla.core.safe(formula = formula, family = family, contrasts = contrasts,  : 
  *** Failed to get good enough initial values. Maybe it is due to something else.

原因是 INLA 23.9.9 与新版 Matrix 1.6.2 不兼容（Matrix 1.6.2 发布很多更新），需要等待新版 Matrix 或 INLA 发布之后，再看。

[XiangyunHuang]

安装 Matrix 旧版本 1.6.-1.1 可以暂时解决此问题

[XiangyunHuang]

目前的开发版 Matrix 1.6-3 和 INLA 23.11.12 也不行，报错如下

Error in validObject(.Object) : 
  invalid class “dsparseModelMatrix” object: superclass "xMatrix" not defined in the environment of the object's class

只有回滚 Matrix 版本或重装 MatrixModels 才行，原因是 hrue/r-inla#85

[github-actions]

🚀 Deployed on https://6554b1a87d54b100a79f85d2--data-analysis-in-action.netlify.app