0

I have weekly data starting from 22May2016 till 30Apr2017. I have data in column backlog, when I create time series data using this column and try to forecast it doesn't capture data from backlog column. Instade take some different number, below is code I am using, can anyone please help me how can I do time series on weekly data 

data <- read.csv("am.csv",header=T,na.strings=c(""))

data <- data[(data$ID %in% c('905')),]

data <- subset(data,select=c(Backlog))

ts <- ts(data$Backlog, frequency=52)

fit <- ets(ts)
plot(forecast(fit))


ID  Date    Backlog
905 1-Jan-17    0.97
905 2-Apr-17    0.95
905 2-Oct-16    0.98
905 3-Jul-16    0.96
905 4-Dec-16    0.96
905 4-Sep-16    0.99
905 5-Feb-17    0.94
905 5-Jun-16    0.98
905 5-Mar-17    0.94
905 6-Nov-16    0.91
905 7-Aug-16    0.95
905 8-Jan-17    0.67
905 9-Apr-17    0.97
905 9-Oct-16    0.71
905 10-Jul-16   0.92
905 11-Dec-16   0.92
905 11-Sep-16   0.93
905 12-Feb-17   0.95
905 12-Jun-16   0.87
905 12-Mar-17   0.88
905 13-Nov-16   1
905 14-Aug-16   0.87
905 15-Jan-17   0.89
905 16-Apr-17   0.99
905 16-Oct-16   0.84
905 17-Jul-16   0.9
905 18-Dec-16   0.83
905 18-Sep-16   0.94
905 19-Feb-17   1
905 19-Jun-16   0.95
905 19-Mar-17   0.97
905 20-Nov-16   0.96
905 21-Aug-16   0.99
905 22-Jan-17   0.92
905 22-May-16   0.99
905 23-Apr-17   0.98
905 23-Oct-16   0.86
905 24-Jul-16   0.91
905 25-Dec-16   0.93
905 25-Sep-16   0.96
905 26-Feb-17   0.98
905 26-Jun-16   0.91
905 26-Mar-17   0.95
905 27-Nov-16   0.92
905 28-Aug-16   0.95
905 29-Jan-17   0.95
905 29-May-16   0.96
905 30-Apr-17   0.94
905 30-Oct-16   0.92
905 31-Jul-16   0.96
user3734568
  • 1,311
  • 2
  • 22
  • 36
  • 1
    Could you give a [reproducible example](http://stackoverflow.com/questions/5963269/how-to-make-a-great-r-reproducible-example/5963610) of `am.csv`? That would make it easier for others to help you. – Jaap May 30 '17 at 11:43
  • @Jaap, As requested I have inserted .csv file in my earlier query – user3734568 May 30 '17 at 11:53
  • @Jaap, think I need to convert column on which I need to perform to vector, then convert it to ts after that it does capture correct data, I have added below script and was able to capture correct data, but further to backlog <- as.vector(Data$Overall.Backlog.Processing.Efficiency.Incident) But further to that forecasting is not working fit <- ets(ts) plot(forecast(fit)) – user3734568 May 31 '17 at 09:23

0 Answers0