Hi, I don't really know if I am missing something here, but when I use Weka for time series forecasting through its graphical frontend, it works nicely and fast, but when I try to do the same programmatically, I get the following error:
I am following this simple example: http://martinbmadsen.dk/weka-time-se...d-values-date/ , and yes, it works perfectly with the wine file, but it seems my file is too big, which is weird since, as I already said, it works great with the Weka frontend. My file has approximately 15000 records and only two data columns. I have also modified the memory parameters of the JVM.
I have attached my arff data file.
Thank you very much for your help!
Code:
Transforming input data...
Exception in thread "main" java.lang.OutOfMemoryError: Java heap space
at weka.classifiers.functions.GaussianProcesses.buildClassifier(GaussianProcesses.java:384)
at weka.classifiers.timeseries.WekaForecaster$SingleTargetForecaster.buildForecaster(WekaForecaster.java:956)
at weka.classifiers.timeseries.WekaForecaster.buildForecaster(WekaForecaster.java:1053)
at TimeSeriesExample.main(TimeSeriesExample.java:48)I have attached my arff data file.
Thank you very much for your help!