Compare classical and Bayesian linear regression

[Shilpa Jose]

Hi,

I have doubt that whether we can compare the Ordinary linear regression and Bayesian regression with INLA to show Bayesian is a better model.

Also when tried running the below mentioned error is showing. Kindly help me with this.

> dataNMMAPS <- read_excel("C:\\Users\\lenovo\\OneDrive\\New folder\\NMMAPSraw.xlsx") > formula <- pm10 ~ 1 + temperature > model.linear<- inla(formula , family = "gaussian" ,data=dataNMMAPS) Error in inla.core(formula = formula, family = family, contrasts = contrasts, : The response for family[1] is not of type 'numeric|list|matrix'; don't know what to do. *** inla.core.safe: inla.program has crashed: rerun to get better initial values. try=1/2 Error in inla.core(formula = formula, family = family, contrasts = contrasts, : The response for family[1] is not of type 'numeric|list|matrix'; don't know what to do. *** inla.core.safe: inla.program has crashed: rerun to get better initial values. try=2/2 Error in inla.core(formula = formula, family = family, contrasts = contrasts, : The response for family[1] is not of type 'numeric|list|matrix'; don't know what to do. Error in inla.core.safe(formula = formula, family = family, contrasts = contrasts, : *** Fail to get good enough initial values. Maybe it is due to something else.