{smcl}
{* June 16, 2006}{...}
{hline}
help for {hi: dpredict}{right: juraj.katriak@iskwien.at}
{hline}

{title:Title}
{p 8 17 2}
{hi: dpredict} {hline 2} Dynamic forecasting  for ARIMA(p,d,q) models
{p_end}

{title:Syntax}

{p 8 17 2}
{cmdab:dpredict} 
[{it:varlist} ]
{cmd:,}
{it:from(#)} 
{it:to(#)} 
{it:periods(#)}
[{it:lvl}]
{cmd::}
{it:command}
{p_end}

{synoptset 26 tabbed}{...}
{synopthdr}
{synoptline}
{synopt: {opt f:rom(#)}}the number of the last observation of the first subsample{p_end}
{synopt:{opt t:o(#)}}the number of the last observation of the last subsample {p_end}
{synopt:{opt p:eriods(#)}}horizon (no. of periods) of the dynamic forecasts{p_end}
{synopt:{opt l:vl}}specifies that the dependent variable of the arima-command is in levels, rather than in differences{p_end}
{synopt:{opt nor:ecursive}}specifies that the subsamples will have a fixed size, rather than grow{p_end}
{synoptline}
{p 4 17 2}
{it:command} must be a valid arima estimation command

{title:Description}

{pstd}{cmd:dpredict} calculates dynamic predictions from the {it:command}, which  must be the {it:{help arima}} command. 
The predictions calculated by {cmd:dpredict} are all out-of-sample predictions, each based on a different subsample. The horizon 
of the predictions (how many periods ahead) is specified by the {it:periods(#)} option. The subsamples, which the 
predictions are based on, change sequentially starting with the subsample delimited by 1/{it:from(#)} up to 1/{it:to(#)}. If the {it:norecursive} option is specified, then the subsamples will have a fixed size of {it:from(#)} observations, i.e. 
the first subsample will be 1/{it:from(#)} and the last ({it:to(#)}-{it:from(#)})/{it:to(#)}.

{pstd}
If the {it:lvl} options is not specified, then the direction-of-change statistic is computed assuming that
the dependent variable of the supplied arima-commmand is in differences corresponding to the 
forecast horizon specified in {it:periods(#)} (i.e. if the forecast horizon is e.g. 3 then the it is assumed that the
dependent variable is of the form of s3.{it:dep_var}). 

{pstd}
If {it:varlist} is supplied then in each step the last observation in the current subsample of the all variables 
in specified in arima (including the dependent) is replaced by the corresponding observation from {it:varlist}. Hence, if {it:varlist} is supplied,
it must contain the same number of variables as the arima command. This allows for the possibility that the last observation of some
variables in the dataset must, in fact, be predicted, because it is not available at the time when making the forecasts.

{title:Saved Results}

{p 4 4 2}
{cmd:dpredict} saves in r(): scalars , r(RMSE), r(MAE) and r(DOC), containing the computed statistcs RMSE, MAE, DOC and matrices , r(Yhat) and r(ResMat), which contain the out-of-sample predictions and residuals respectively. Additionaly a matrix , r(DOCmat), with elements 1 (if the direction of change was predicted correctly) and 0 (if not) is saved.
{p_end}

{title:Remarks}

{p 4 4 2}
The statistisc computed along with the predictions are: RMSE - root mean square error, MAE - mean absolute errror and DOE - direction of change (gives the proportion of correctly predicted values of {it:dep_var})
{p_end}

{title:Examples}

{p 4 8 2}
{cmd:. use http://www.stata-press.com/data/r9/wpi1}

{p 4 8 2}
{cmd:. gen wpi_df = d.wpi}

{p 4 8 2}
{cmd:. dpredict , f(80) t(124) p(1) : arima wpi_df  , arima(1,0,1)}

{title:Author}

{p 4 4 2}
{browse "mailto:juraj.katriak@iskwien.at":J. Katriak}, ISK Vienna, juraj.katriak@iskwien.at
{p_end}

{title:Also see}

{p 4 13 2}
Online: help for {help arima}, {help arima_postestimation}
{p_end}