Reputation: 1
"I am trying to use the panelAR package for a Panel Regression model with AR(1) correction and panel-corrected standard errors. The package has several arguments however when I run the model I receive an error "Error: The time variable must be defined as an integer". Even though the time variable "year" is st as an integer".
It's probably a simple fix but would appreciate any helpful advice. Thanks
data$year <-as.integer(data$year) data$area <-as.integer(data$area) str(data)
corrmodel <- panelAR(crmrte ~density+pctymle+pctclaims+rdcrte+antrte+pctleavFE+
pctleavHE+pctleavEMP+pctleavUn, data= pdata, panelVar = "area",
timeVar = "year", autoCorr = "ar1", panelCorrMethod = "pcse",
rhotype = "breg",bound.rho = FALSE,rho.na.rm = FALSE,
panel.weight = "t-1",dof.correction = FALSE,complete.case = FALSE,
seq.times = FALSE, singular.ok=TRUE)
Upvotes: 0
Views: 185
Reputation: 11
For posterity: Apparently panelAR
doesn't like tibbles/tbl_df
. Applying as.data.frame()
to my input data set solved the issue as noted here: https://github.com/kkashin/panelAR/issues/2
Upvotes: 1