similar to: problems with residuals of an arima model

Dear list and Martin, I'm testing different approaches to fit an electricity demand time series and come upon the fracdiff package (v 1.3-1) for fitting fractional ARIMA models. The following questions are motivated by this package. 1. Despite having a help page, the residuals and fitted functions don't seem to have implementation, or did i miss something obvious? Alternatively, having a

Hi I have fitted an arima model to a monthly series of length 113. I have use fit <- arima(x.ts, order = c(1,0,0),seas = list(order = c(1,2,1),12, method = "CSS") I'm a little confused because the series of residual has the first 46 values equal to 0. From there the residuals are approximately while noise, numbers have a mean close to 0 and a variance value. What could be

Hi I am fitting an arima model to some time series X. When I was comparing the fitted values of the model to the true time series I realized that the true time series lags one time step behind the fitted values of the arima model. And this is the case for any model. When I did a simple linear regression using lm to check, I also find the same results, that the true series lags behind the

Hi I'm trying to fit arima models with the arima() function and I have two questions. ###### ##1. ## ###### I have n observations for my time series. Now, no matter what arima(p,d,q)- model I fit, I always get n residuals. How is that possible? For example: If I try this out myself on an AR(1) and calculate the fitted values from the estimated coefficients I can calculate n-1 residuals.

Hi, Although my doubt is pretty,as i m not from stats background i am not sure how to proceed on this. Currently i am doing a forecasting.I used ARIMA to forecast and time series was volatile i used garchFit for residuals. How to use the output of Garch to correct the forecasted values from ARIMA. Here is my code: ###delta is the data fit<-arima(delta,order=c(2,,0,1)) fit.res <-

I am trying to find a way to obtain the fitted values for a model fit using arima() in the ts package. I came across a suggestion in the mailing list archive that these values can be simply calculated as: model<-arima(t, order = c(1,1,0)); fitted<-t-model$residuals; But, the help file for arima() in the ts package describes the residuals values returned as being "standardized

Dear colleagues, I am relatively new to R and time series and so I am experiencing difficulties in interpreting the output of "arima" in MA models (but not in AR models). I cannot make sense of the 1st innovations returned by "arima". In an AR(1) model I expect data[t]=phi1*data[t-1]+a[t] and in a MA(1) model data[t]=a[t]+theta1*a[t-1]. My interpretation from R-help is

Dear all, I'm looking for how can I input a linear model with an arma model,like log(y) = 8.95756 + 0.0346414^t - 0.1*t^2 + ut ut=-0.296ut-1+at-0.68at-1 where log(y) is qudratic function ,for the time series trend, and get then get the residuals from the first function. " obersvations value - the fit value = ut" and fit an ARIMA(1,1,1) model for ut.

Dear All, I decided to use an AR(1) model for the residual series and trying to find the p-value for each parameter by using arima command in R, but i never find it from the output. The output gives me the parameter and mean's value, its standard error, estaimated variance, AIC and loglikehood, but no p-value is given. What i did was, i assign the residuals into the 'residual'

Hello, Does anybody know if the function arima with an external regressor (xreg) applies the auto correlation on the dependant variable or on the residuals. In comparison with SAS (proc autoreg), it seems that the auto correlation applies on the residuals but i'd like to have the confirmation. I want to estimate: Y[t] = a[1]*X[t] + a[2] + E[t] with E[t]=b[1]*E[t-1] Should I use : arima(Y,

Good afternoon! I'm trying to model a time series on the following data, which represent a monthly consumption of juices: >x<-scan() 1: 2859 3613 3930 5193 4523 3226 4280 3436 3235 3379 3517 6022 13: 4465 4604 5441 6575 6092 6607 6390 6150 6488 5912 6228 10196 25: 7612 7270 8617 9535 8449 8520 9148 8077 7824 7991 7660 12130 37: 9135 9512 9631 12642

I've been trying to get comfy with arima() and associated functions in the ts() package. I'm thinking seriously about using this package, and R generally, in a 4th year intro time series course that I'm teaching this autumn. I have a couple of questions about arima: (1) The help file says that residuals component of the value returned by arima() consists of the

two ARIMA models, both have several bars signicant in ACF and PACF plots of their residuals, but when run Ljung Box tests, both don't show any significant correlations... however, one model has p-value that is larger than the other model, based on the p-values, can I say the model with larger p-values should be better than the model with smaller p-values? [[alternative HTML version

Hi all, Forecasting from an arima model is easy with predict. But I can't manage to backcast : invent data from the model before the begining of the sample. The theory is easy : take your parameters, reverse your data, forecast, and then reverse the forecast I've tried to adapt the predict function to do that (i'm not sure that the statistical procedure is fine (with the residuals),

Hi Everyone, I'm having a problem with arima.sim. Namely specifying inital values for the series. If I generate a random walk > vs = rnorm(100,0,1) > xs = cumsum(vs) and fit an ARIMA(1,0,0) to it > xarima = arima(xs,order=c(1,0,0)) > xarima Call: arima(x = xs, order = c(1, 0, 0)) Coefficients: ar1 intercept 0.9895 8.6341 s.e. 0.0106 6.1869 I should

Hi This is my first post to this group, so apologies in advance if I get it wrong. I would like to know how the prediction for arima models works in R. I have a time series to which I fit an arima model, of varying AR and MA orders. I then use the predict function to project it forward. I have also written my own function to perform the prediction, but it gives different answers to Arima.predict

Hi, I have the following code from Splus that I'd like to migrate to R. So far, the only problem is the arima.filt function. This function allows me to filter an existing time-series through a previously estimated arima model, and obtain the residuals for further use. Here's the Splus code: # x is the estimation time series, new.infl is a timeseries that contains new information # a.mle

Hi all, I want to calculate the residuals of an ARIMA fit to a time series using the same calculation method as in arima() in the ts package, BUT I want to be able to specify fixed values for ALL the parameters in the ARIMA model. The "fixed = ..." argument in the arima() command appears to require at least one parameter to be variable. Does anyone know a way around this?

Hi, I've fitted a GARCH(1,1) for the residuals of my time serie (X). X is an ARMA(1,1) process. Now I want to do a n-step forecast for X, knowing these processes. How can I do this? I know that there's a command: predict() for ARIMA processes and so on, but what about GARCH? I've got: arma=arima(x, order=c(1,0,1)) (...) garch11<-garch(residuals(x),order = c(1, 1))

Full_Name: Allan McRae Version: 1.6.0 OS: Win 2000 P Submission from: (NULL) (129.215.190.229) When using predict.Arima in library ts(), it appears differencing is only accounted for in the first step of prediction and so any trend is not apparent in the predictions. The example shows the difference between the predictions of an arima(1,1,1) model and the backtransformed predictions of an