0
$\begingroup$

I have a time series for daily records of concentration of Total Phosphorus (TP) in a riiver and the water flow in the year of 2014. While I have complete data for the flow variable, I have missing values for the TP variable. I am trying to impute the missing values of the TP variable based on an AR(1) model for TP with the flow variable as an external regressor. The model is programmed in R as: ts1=arima(TP, order = c(1,1,0), xreg=F)

I came across imputeTS package on R and tried the command na.kalman. I want my model ts to be used hence I used the argument model in the command as follows:

na.kalman(TP, model = "ts")

This method does not seem to be working as I get the error

Error in na.kalman(tp1, model = "ts1_") : Parameter model has either to be "StructTS"/"auto.arima" or a user supplied model in form of a list with at least components T, Z, h , V, a, P, Pn specified

What did I do wrong here? How can I use the command na.kalman and to make it make use my ARIMA model?

$\endgroup$
4
$\begingroup$

You need to pass it the model and not a string:

ts1=arima(TP, order = c(1,1,0), xreg=F)

na.kalman(TP, model = ts1$model)

$\endgroup$
0

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.