Search code examples
requationbayesianlog-likelihood

Converting likelihood equation into R code

I'm working on coding GP model in R and calculating its likelihood function.

I'm having problem to convert the equation below into R code: enter image description here

I used the code below but I keep getting error message :

Error in (t(zlt - olt)) * (zlt - olt) : non-conformable arrays

my code is:

ConditionalLikelihood <- (-(N/2)) * (log(sig2e)) -(1/(2*(sig2e)))*(t(zlt -olt)) * (zlt- olt) -(sum(T*r)/2) *(log (sig2eta * SEta)) (-1/(2*(sig2e))) * (t(zlt- xb)) * 1/SEta *(zlt- xb)

I think it's related to the 2 sum signs. I couldn't find an example shows how to deal with 2 sum signs and follow their structure.

I tried to debug it part by part to check where I went wrong and every time I get to this part , I get the error:

(t(zlt -olt)) * (zlt- olt)

Any advice will be much appreciated.

Solution

  • Try %*% for matrix multiplication instead of *.

    (t(zlt -olt)) %*% (zlt- olt)