做緊個Time Series既project....
我想知如果揀左兩個variable做完unit root test知道兩個都係I(1) : non-stationary
之後做個Engle-Granger cointegration test (run unit root test on residuals)又發現佢地唔cointegrated既話
仲可唔可以做其他model ? e.g. VAR, Granger causality test.... ?
另外其實我唔係好明Engle-Granger cointegration同Granger causality test既分別.....
前者係睇下兩個variables係long run度有無meaningful relationship...後者係check有無one-way / two-way causality ??
真係做到一頭煙....希望有高人指點.....
