Charles Thuo
2015-Jul-01 04:22 UTC
[R] How to run a glm using one factor in a dichotomous categorical variable
am running a glm to model the number of accidents that result from diesel or petrol cars. How can i run a glm that on only one factor i.e. diesel only. require(CASdatasets) data(freMTPL2freq) CONTRACTS<- data.frame(freMTPL2freq$IDpol,freMTPL2freq$ClaimNb,freMTPL2freq$Exposure,freMTPL2freq$VehPower,freMTPL2freq$VehAge,freMTPL2freq$DrivAge, freMTPL2freq$VehBrand,freMTPL2freq$VehGas,freMTPL2freq$Region,freMTPL2freq$Density) colnames(CONTRACTS)<- c("IDpol","ClaimNb","Exposure","VehPower","VehAge","DrivAge","VehBrand","VehGas","Region","Density") CONTRACTS.f<-CONTRACTS vY<- CONTRACTS.f$ClaimNb head(vY) vE<- CONTRACTS.f$Exposure head(vE) X1<-CONTRACTS.f$VehGas;head(X1) tapply(vY,X1,sum) tapply(vY,X1,sum)/tapply(vE,X1,sum) df<- data.frame(vY,vE,X1);head(df);tail(df) regpoislog<- glm(vY~0+X1+offset(log(vE)),data=df,family=poisson(link="log")) coef(regpoislog) X1Diesel X1Regular -2.327388 -2.267530 Kindly assist [[alternative HTML version deleted]]