bpfaff / vars

Multivariate Time Series Models: VAR, SVAR and SVEC
40 stars 23 forks source link

R-Squared statistic does not seem to be correct when using restricted models #15

Open pgg1309 opened 2 years ago

pgg1309 commented 2 years ago

Hello, I noticed very different results for R-Squared when printing the results of a restrict model. Looking at the examples given in the package

library(vars) data(Canada) var.2c <- VAR(Canada, p = 2, type = "const") summary(var.2c) var.2c.restrict <- restrict(var.2c, method = "ser") summary(var.2c.restrict)

It does not seem that the summary(var.2c.restrict) gives the correct R-Squared statistics. Is that a but or am I missing something?

Thanks and congrats on this very useful package!