forecasting

I have a neural network created using PyBrain and designed to forecast time series. I am using the sequential dataset function, and trying to use a sliding window of 5 previous values to predict the 6th. One of my problems is that I can't figure out how to create the required dataset by appending the 5 previous values to the inputs and the 6th as an output. I am also unsure of how exactly to forecast values in the series once the network is trained. Posting my code below: from pybrain.datasets import SupervisedDataSet from pybrain.datasets import SequentialDataSet from pybrain.tools.shortcuts

I was looking through a tutorial by Rob J Hyndman, here's the link for the tutorial , my question is how can i add a fitted line to the forecast plot eg; library(forecast) library(ggplot2) fc <- forecast(fdeaths) autoplot(fc) I now need to add fitted(fc) to the plot above, how do I do it? Here an other solution without additional packages: fit <- data.frame(data=as.matrix(fitted(fc)), date=time(fitted(fc))) autoplot(fc) + geom_line(data = fit,aes(date,data), col = "red") You transform your ts into a data frame where you can use normal ggplot command. You can then just add the line. I realized

I have two years of monthly data but stl() seems to need a minimum of two years and one month . Here are two simple examples: Example 1 - returns Error in stl(x, "periodic") : series is not periodic or has less than two periods dat_24 <- cumsum(rnorm(24)) x_24 <- ts(dat_24, frequency = 12) stl(x_24, "periodic") Example 2 - returns forecast as expected dat_25 <- cumsum(rnorm(25)) x_25 <- ts(dat_25, frequency = 12) stl(x_25, "periodic") Shouldn't I be able to get a forecast with only 24 numbers with frequency = 12? It holds true for all kind of periodic series, be it a weekly or yearly or any

Good day I read on one of the posts here that "the function forecast::plot.forecast is not designed to be used with axis.Date or axis.POSIXct (which are not used in the package forecast)." This can be seen here: custom axis labels plotting a forecast in R Nevertheless, they managed to use the forecast package and some code to get the correct axis labels. However, this example is for quarterly data. Also, this example here using 'as.POSIXlt' is for weekly data: Forecasting time series data I've tried playing with the code but I can't get it to work for monthly data. So my axis labels are still

I have a hierarchical time series, the bottom level series of which all exhibit intermittent demand. It seems advantageous to use Hyndman's HTS package for optimal combination within the hierarchy. It also seems advantageous to use Kourentzes' MAPA package for multiple aggregation prediction of the intermittent demand. In essence, I want to do something like: forecast(my_hts, method='comb', fmethod='MAPA') However, it is unclear to me if / how I can combine the two, since forecast.gts() only accepts fmethod=c("ets", "arima", "rw"). Is there a clever way to pass different forecasting methods to

I'm trying to wrap my head around ARIMA forecasting using Python and Statsmodels. Specifically, for the ARIMA algorithm to work, the data needs to be made stationary via differencing (or similar method). The question is: How does one invert the differencing after the residual forecast has been made to get back to a forecast including the trend and seasonality that was differenced out? (I saw a similar question here but alas, no answers have been posted.) Here's what I've done so far (based on the example in the last chapter of Mastering Python Data Analysis , Magnus Vilhelm Persson; Luiz