ããæå¼ä½ è¦å»ºæ¨¡çåºå,å设æ¯x,ç¹è¿ä¸ªåéçªå£å·¥å
·æ éçview-correlogram.
ããè¿éæå 个åæ°ï¼level=0,表示对ååºåä½å¾,1st difference=1表示对ä¸é¶å·®åä½å¾,2nd表示对äºé¶å·®åä½å¾,lags表示æ大æ»åé¶æ°.使ç¨é»è®¤åæ°å°±å¯ä»¥.ææ¶åå¯è½ä¼åºç°near singular matrixçé误,ä½ å¯ä»¥éæè°æ´lagsçåå¼,ç´å°OKå°±è¡.
ããæå®,çå°ä¸¤ä¸ªå¾,autocorrelationèªç¸å
³å¾,partical correlationåèªç¸å
³å¾,å¾ä¸æ
æ¾èæ§æ£éªç临çå¼ç线.æä¹ç¨èªç¸å
³å¾ååèªç¸å
³å¾åå«å¤æmaåarçæ»åé¶æ°,ç¸ä¿¡ä½ æ¯ç¥éçå§.
ãã好äº,åè®¾ä½ æ ¹æ®è¿ä¸¤ä¸ªå¾å¤æåºçmaãaræ»åé¶æ°åå«æ¯q=2,p=3
ããæ以è¦å»ºç«ç模åæ¯ar(2)ma(3)
ãã主çªå£çå·¥å
·æ é,注ææ¯ä¸»çªå£å¦,ç¹å»quick-estimate equation,å¨éé¢è¾å
¥x ar(1) ar(2) ma(1) ma(2) ma(3),å
¶ä»åæ°é»è®¤,OKå°±å¯ä»¥çå°åºæ¬ç模åäº,注æä¸é¢è¾å
¥çåéä¹é´æ¯ç©ºæ ¼,没æåé符å·.
ããå¦æpåqçåå¼ä¸æç¡®,å¯ä»¥å¤å°è¯å 个påqçå¯è½ç»å,ççç¸å
³æ£éªçæ¾èæ§,å
³é®æ¯è¾ç»æä¸çAICåsc,è¶å°è¶å¥½.
ãã麻ç¦çå°æ¹æ¯ï¼
ããè¦æ¯ä½ çåºåä¸å¹³ç¨³,éè¦å»ºç«arima模å,è¿æ¶å°±è¦ç1st differeceçè³2nd differenceçå¾å½¢.
ããè¦æ¯æ¶å12é¶æ»åèªç¸å
³æåèªç¸å
³æ¾è,å°±è¦å¨ç¨sarima模åäº,åèµ·æ¥å®¹æ,å¯ä¸å®¹æ讲æ¸
æ¥å¦.
ããä½ åªè¦æ±arma模å,估计æ¯ä¸éè¦ç¨arimaåsarima模åçå§.