Hi Sportspimp here, turns out that neuralnet no longer uses the compute command, just predict. I thought it worked but error message I thought I could research the fix but no luck. Back for more help

Model is labeled OU, using the command you gave me with predict instead of compute I get this

> OUPred<- predict(OU, new_data)$net.result, %>% sapply(denormalize)

Error: unexpected ',' in "OUPred<- predict(OU, new_data)$net.result,"

Can you figure out what "," is the problem, 

  • No luck > OUPred <- predict(OU, new_data)$net.result %>% sapply(denormalize) Error in predict(OU, new_data)$net.result : $ operator is invalid for atomic vectors

  • > OUPred <- predict(OU, new_data_norm)$net.result %>% sapply(denormalize) Error: object 'new_data_norm' not found and class is nn

Answer

Answers can only be viewed under the following conditions:
  1. The questioner was satisfied with and accepted the answer, or
  2. The answer was evaluated as being 100% correct by the judge.
View the answer
Kav10 Kav10
2.1K
  • Kav10 Kav10
    0

    Predict() directly returns the predictions as a matrix (unlike compute(), which returns a list containing $net.result), so you don’t need $net.result.

  • Kav10 Kav10
    0

    Try removing $net.results.

  • Kav10 Kav10
    0

    It is hard to read code in comments. Please add your full code to the question so I can better assess where the issue may be.

  • I have to step away for an hour, I will respond as soon as I return. Thank you for your time and patience. I will return as fast as I can.

  • Im back, asking new question to allow me to insert code in a legible way

The answer is accepted.
Join Matchmaticians Affiliate Marketing Program to earn up to a 50% commission on every question that your affiliated users ask or answer.