I am looking for a cointegration function that can execute cointegration test faster than functions available in urca library.
Here is my test data from timeSeries package
library(timeSeries)
data = LPP2005REC
test_dt = data.table(data$MPI, data$SBI)
Here are the performance results of cointegration test using urca library.
library(urca)
microbenchmark::microbenchmark(
ca.po.pu = summary(urca::ca.po(test_dt, type = "Pu", demean = "trend", lag = "short"))@testreg$coefficients[2, 1],
ca.jo = urca::ca.jo(test_dt, type = "trace", ecdet = "trend", spec = "longrun", K = 2)@teststat[1],
times = 5)
Unit: milliseconds
expr min lq mean median uq max neval cld
ca.po.pu 9.105962 9.509942 10.777335 10.651319 11.219605 13.399847 5 b
ca.jo 3.646501 3.959293 4.581624 4.926513 5.043406 5.332409 5 a
I will appreciate it if someone can suggest a faster cointegration test function.