0

I did forecasting by mlp through nnfor package, i am satisfied with my results but i need plot of mlp that should have neurons with weights.

fit.mlp <- mlp(infl.train,hd = 3, hd.max =NULL, difforder=NULL,det.type = "bin",
               allow.det.season = TRUE,outplot = TRUE,reps = 100,
               comb = "median")

plot(fit.mlp)

By simple plot fucntion i got the following figure:

enter image description here

it just gives the mlp nodes but not weights on these nodes please suggest me some command which can be used for the purpose of obtaining the neurons with weights in nnfor package of r while forecasting a univariate series.

Chris
  • 3,836
  • 1
  • 16
  • 34
Rizwan Ali
  • 1
  • 5
  • 1
    Welcome to SO! While you've included the code you've used to make the plot, adding a sample of your data will allow others to replicate your problem and make it more likely you will get a response. See [how-to-make-a-great-r-reproducible-example](https://stackoverflow.com/questions/5963269/how-to-make-a-great-r-reproducible-example) – Chris Jul 17 '19 at 19:42
  • @Chris thanks for the suggestion but question is very simple if someone have used nnfor package then they can answer it, there is no need of data or further codes. The codes that i provided are just to understand the question otherwise they are also not required. – Rizwan Ali Jul 18 '19 at 14:01
  • Maybe so but providing some data to play with and a clear description of your expected result might enable people *without* experience of this package to help you – Chris Jul 18 '19 at 14:39

0 Answers0