Forum: help

RE: translogEst with more than two independent variables [ Reply ]
By: Jorge Viana on 2015-07-22 19:28

Dear Arne,

Here is the reproducible example

####################################
# Using translogRayEst()
data(germanFarms)
germanFarms$qCrop = germanFarms$vCrop / germanFarms$pOutput
germanFarms$qAnimal = germanFarms$vAnimal / germanFarms$pOutput
germanFarms$qVarInput = germanFarms$vVarInput / germanFarms$pVarInput
y.translog.ray = translogRayEst(yNames = c("qCrop", "qAnimal"), xNames = c( "qLabor", "land", "qVarInput"),
data = germanFarms )
summary(y.translog.ray)

# Using translogEst()

germanFarms$distance = sqrt(germanFarms$qCrop^2 + germanFarms$qAnimal^2)
germanFarms$theta = acos(germanFarms$qCrop/germanFarms$distance)
y.translog.ray$theta - germanFarms$theta #checking if the thetas are equal
y.translog.ray$distance - germanFarms$distance #checking if the distamces are equal

y.translog = translogEst(yName = "distance", xNames = c("qLabor", "land", "qVarInput", "theta"),
data = germanFarms)
summary(y.translog)
############################

Thanks again,
Jorge

RE: translogEst with more than two independent variables [ Reply ] By: Arne Henningsen on 2015-07-21 15:44	[forum:42383]
Dear Jorge Please provide a reproducible example. Best regards, Arne

translogEst with more than two independent variables [ Reply ]
By: Jorge Viana on 2015-07-21 13:46

[forum:42380]

Since translogRayEst () can`t estimate a model with more than two independent variables I tried to use translogEst(), with the norm of my two-dimensional vector as independent variable and using "theta" as an aditional dependent variable to test if I could extend thi resioning two more than two independent variables. However, I got parameters estimation quite similar for the original dependent variables and very diferent for the "theta" variable, comparing with results obtained using translogRayEst (). Does anyone can explain why is that? I thought they would excatly the same.

Tkanks
Jorge