I am trying to run a first order VAR that involves seven state variables. I keep getting the error below:
US_VAR=VAR(VARR, type="const", lag.max=1)
Error in lm.fit(x = ys.lagged, y = yendog) : NA/NaN/Inf in 'x'
I'm not sure what is wrong, there are no NAs in my dataset. Any ideas on what the problem is?
This question is rather ambiguous as the command VAR is multiply defined; which VAR? Judging from the syntax, I am guessing {vars}.
The error message says that x, which has the word lagged in the name and that's often a hint that it was time shifted, contains values that are either NA [the first or last observation is often missing because of the implied time shift], NaN [not a number] or infinite, though possibly negative infinite especially if one has not been exceedingly careful about the fact that logs are only defined on non-negatives. Absent additional information, this is going to be really hard to diagnose. The command works; the reprex in the help file works with either the {fable} or {vars}.
By the way, I am not sure that RStudio IDE is the proper topic heading for this as no version of VAR is RStudio or their IDE.
This topic was automatically closed 21 days after the last reply. New replies are no longer allowed.
If you have a query related to it or one of the replies, start a new topic and refer back with a link.