Dynare forums

Hi all,

In Bayes estimation, I want to match my model to several observed variables.

Variable Y, C, I have 100 observations, while variable B has only 70 observations. Can I use all the observations of Y, C, I, B ? Or do they have to be the same length ?

If balanced length must be required. I have to use Y, C, I with 100 observations or Y, C, I, B with only 70 observations... I wander which way performs better?

Thanks!

Dynare's Kalman filter supports missing observations. Therefore you can use all 100 observations. You simply need to code the missing observations as NaN in the dataset.

jpfeifer wrote:Dynare's Kalman filter supports missing observations. Therefore you can use all 100 observations. You simply need to code the missing observations as NaN in the dataset.

Thank u Professor!