How to use the nnts() function
nnts() fits a single-hidden-layer neural network to a time series. The
calling function is
nnts(data,lags,size,retry=1,maxit=2000, trace=F,nntrace=F,...)
- data
- vector containing the data.
- lags
- vector containing the lags.
- size
- number of units in the hidden layer.
- retry
- number of starts of the fitting algorithm. Default is one.
- maxit
- maximum number of iterations. Default 2000.
- trace
- show the result of each start. Default False.
- nntrace
- switch for tracing optimization. Default False.
- ...
- remaining arguments to be passed to nnet().
The returned object has the same internal structure as an nnet object,
but also has
- x
- the lagged inputs returned as a matrix.
- y
- the output.
- lags
- the lags used.
This is a wrapper function to call nnet() for time series data.
Last modified: Sat Aug 18 10:50:54 PDT 2012