用ARIMA/指数平滑法在R

我已经编写了一些代码来对索赔日期进行排序，每月/每年对它们进行计数，并尝试使用ARIMA/指数平滑参数对它们进行预测

见索赔清单：

            2012-01-31   82
            2012-02-29   65     
            2012-03-31   64
            2012-04-30   73
            2012-05-31   71
            2012-06-30   79
            2012-07-31   72
            2012-08-31   82
            2012-09-29   64
            2012-10-31   72
            2012-11-30   63
            2012-12-31   80
            2013-01-31   67             
            2013-02-27   65
            2013-03-31   84
            2013-04-30   68
            2013-05-31   68
            2013-06-29   66
            2013-07-30   64
            2013-08-31   69
            2013-09-29   66
            2013-10-31   65
            2013-11-30   56
            2013-12-31   76
            2014-01-31   75
            2014-02-28   58
            2014-03-29   80
            2014-04-30   76
            2014-05-31   80
            2014-06-28   68
            2014-07-31   82
            2014-08-30   79
            2014-09-27   60
            2014-10-31   85
            2014-11-30   60
            2014-12-31   76
            2015-01-31   75
            2015-02-28   84
            2015-03-31   77
            2015-04-30   79
            2015-05-30   91
            2015-06-30   82
            2015-07-31   98
            2015-08-31   65
            2015-09-30   77
            2015-10-31  115
            2015-11-30   79
            2015-12-31   80
            2016-01-30   91
            2016-02-29  105
            2016-03-31   77
            2016-04-30  107
            2016-05-31   85
            2016-06-30   89
            2016-07-30  112
            2016-08-31   88
            2016-09-30   90
            2016-10-30   79
            2016-11-30   85
            2016-12-31   66

我的代码面临的问题是，我得到的是一个平均预测值，而不是我想要的拟合数据，类似于以下示例：

请参见R代码：

对索赔日期进行排序并计数

        library(forecast)
        library(ggplot2)
        library(xts)
        library(reshape2)
        library(zoo)
        library(lubridate)

        data = read.csv('Claims1.csv')
        data$DISABILITYDATE <- as.Date(data$DISABILITYDATE, "%m/%d/%Y")
        data

        str(data)
        as.Date(data[,1])
        xts(x=data[,-1], order.by = data[,1])

        data = read.csv('Claims1.csv')
        data$DISABILITYDATE <- as.Date  (data$DISABILITYDATE, "%m/%d/%Y")
        df <- xts(rep(1,length(data$DISABILITYDATE)),order.by=data$DISABILITYDATE)
        df1 <- apply.monthly(df,function(x) length(x))
        df1
        t(df1)

        str(df1)
        df2 <- data.frame(df1=c("Jan 2012","Feb 2012","Mar 2012","Apr 2012","May 2012","Jun 2012","Jul 2012","Aug 2012","Sep 2012","Oct 2012","Nov 2012","Dec 2012","Jan 2013","Feb 2013","Mar 2013","Apr 2013","May 2013","Jun 2013","Jul 2013","Aug 2013","Sep 2013","Oct 2013","Nov 2013","Dec 2013","Jan 2014","Feb 2014","Mar 2014","Apr 2014","May 2014","Jun 2014","Jul 2014","Aug 2014","Sep 2014","Oct 2014","Nov 2014","Dec 2014","Jan 2015","Feb 2015","Mar 2015","Apr 2015","May 2015","Jun 2015","Jul 2015","Aug 2015","Sep 2015","Oct 2015","Nov 2015","Dec 2015","Jan 2016","Feb 2016","Mar 2016","Apr 2016","May 2016","Jun 2016","Jul 2016","Aug 2016","Sep 2016","Oct 2016","Nov 2016","Dec 2016"),score=c(df1))
        df2
        t(df2)
        df2[-1]

我的输出是：

期望输出的示例

---

是什么让你认为你的数据是季节性的？仅仅因为你有月度数据并不意味着季节性模型最适合你。是什么让你认为你的数据是季节性的？仅仅因为你有月度数据并不意味着季节性模型最适合它。

        library(forecast)
        x.ts <- as.ts(df2[2])
        x.ts
        x.ets <- ets(x.ts)
        x.ets
        x.fore <- forecast(x.ets$fitted, h=12)
        x.fore

        x <- ts(df2[2], start = 2012, frequency = 12)
        plot(forecast(ets(x), 24))
        x

        plot(forecast(x, h=12))
        date1 <- ymd("2012-01-01","2013-01-01","2014-01-01","2015-01-01","2016-01-01","2017-01-01")
        abline(v=decimal_date(date1), col="blue")

            ARIMAfit = auto.arima(x, approximation=FALSE,trace=FALSE)
            summary(ARIMAfit)
            plot(ARIMAfit)

            pred = predict(ARIMAfit, n.ahead = 48)
            round(as.numeric(pred$fitted,0))
            pred

            library(TSPred)             
            plotarimapred(pred$pred,x, xlim=c(2012, 2020), range.percent = 0.05)