1

I read in a csv file into a data frame using:

dataxlsx <- read.csv(file.choose(), header = T)

The data frame looks like:

Year  Month Period X410     X430     X431
2005  1       1    3467748  4434879  1345638
2005  2       2    3626338  4311150  1167523
  .   .       .      .        .        . 
2015  7       127  2374105  1514540  1399804

I am trying to run my function that I created called HWplot to forecast the inputted data and run the forecast, as well as output a plot for the forecast.

I used packages ggplot2, tseries, forecast.

HWplot <- function(dataxlsx,  n.ahead=12,  CI=.95,  error.ribbon='green', line.size=1) {

hw_object<-HoltWinters(dataxlsx)

forecast<-predict(hw_object,  n.ahead=24,  prediction.interval=T,  level=0.95)


for_values<-data.frame(time=round(time(forecast),  3),  value_forecast=as.data.frame(forecast)$fit,  dev=as.data.frame(forecast)$upr-as.data.frame(forecast)$fit)

fitted_values<-data.frame(time=round(time(hw_object$fitted),  3),  value_fitted=as.data.frame(hw_object$fitted)$xhat)

actual_values<-data.frame(time=round(time(hw_object$x),  3),  Actual=c(hw_object$x))


graphset<-merge(actual_values,  fitted_values,  by='time',  all=TRUE)
graphset<-merge(graphset,  for_values,  all=TRUE,  by='time')
graphset[is.na(graphset$dev),  ]$dev<-0

graphset$Fitted<-c(rep(NA,  NROW(graphset)-(NROW(for_values) + NROW(fitted_values))),  fitted_values$value_fitted,  for_values$value_forecast)


graphset.melt<-melt(graphset[, c('time', 'Actual', 'Fitted')], id='time')

p<-ggplot(graphset.melt,  aes(x=time,  y=value)) + geom_ribbon(data=graphset, aes(x=time, y=Fitted, ymin=Fitted-dev,  ymax=Fitted + dev),  alpha=.2,  fill=error.ribbon) + geom_line(aes(colour=variable), size=line.size) + geom_vline(x=max(actual_values$time),  lty=2) + xlab('Time') + ylab('Value') + theme(legend.position='bottom') + scale_colour_hue('')
return(p)

}

The problem I am facing is that I cannot split this data frame in order to apply the HWplot function to the separate columns of the data (forecast X410, X430, X431, etc). I will be working with data frames that have a different number of X### codes in the columns, so I will need the R Script to incorporate a dynamic number of columns as well.

The end game is to run these forecasts from the different columns of the data frame and output the forecast and graph to an excel workbook with the name of the column as each sheet name.

Side note: The HWplot function works when there is a data frame with only one column of metrics, however is not working with multiple columns of metrics.

Everything I have tried with the apply family of functions isn't working nor are the split functions.

Hope this makes sense - if anyone needs clarification, please let me know.

Jaap
  • 81,064
  • 34
  • 182
  • 193
Michael Donofrio
  • 11
  • 3
  • 1
    Welcome to SO. Would you please be a little more focused in what exactly your problem is, and what the output should look like, and read about [how to make a great R reproducible example](http://stackoverflow.com/questions/5963269/how-to-make-a-great-r-reproducible-example). – C8H10N4O2 Sep 08 '15 at 19:29
  • Thank you! Basically, I am trying to read in a .csv file into R with dates listed and multiple columns of data that need to be forecasted based on the HWplot function I wrote above. My problem is that I don't know how to use the apply functions or loops to read the .csv file, forecast each column individually, and output the forecast data and plot to an excel workbook, where each sheet in the workbook is titled with the column name that was forecasted. I hope this clarifies. – Michael Donofrio Sep 08 '15 at 20:09
  • So you've already read the .csv file -- you can edit that out. Identify what your first question is -- I think it is forecasting the columns individually -- and then edit out the rest for this post. (One question per post.) Once you get that down, if you still need help with the next step, ask another question. I am telling you this to try to help you, because nobody will write your entire program for you in a single question – C8H10N4O2 Sep 09 '15 at 01:59

1 Answers1

1

The "Batch Forecasting in R" blog post in R-bloggers does an excellent job of explaining how to do this.

http://www.r-bloggers.com/batch-forecasting-in-r-2/

Eric Ness
  • 10,119
  • 15
  • 48
  • 51