R 语言 ARMA 建模

stats::arima() 可以用最大似然方法估计 AR、MA、ARMA 和 ARIMA 模型，需要人为指定 (?, ?, ?) 值。如：

armares <- arima( 100 + xarma42, order = c(4,0,2) ) armares ## ## Call: ## arima(x = 100 + xarma42, order = c(4, 0, 2)) ## ## Coefficients: ## ar1 ar2 ar3 ar4 ma1 ma2 intercept ## -0.6324 -1.0668 -0.4163 -0.4469 0.3191 -0.6423 99.9989 ## s.e. 0.1195 0.1384 0.1350 0.1080 0.1237 0.1267 0.0371 ## ## sigma^2 estimated as 3.582: log likelihood = -209.2, aic = 434.4  

R 语言 ARMA 建模最先出现在Python成神之路。