CODEWITHSUNDEEP
r
fyselle
fyselle

Reputation: 25

run 100 model at once, instead of manually one by one

i created this table. 

    dat <- matrix(rbinom(6*100, 6, 0.5), ncol=6)

then use one row from at the table to fill the 'size' statement to train the model. 

    nn.sizes <- dat[1,]
    sample.number <- length(dat2[[1]][,1])
    nns <- lapply(1:length(dat2), function(i){nn <- nnet(dat2[[i]][1:(sample.number),], dat3[[i]][(twindow+1):(length(dat3[[i]]))],size=nn.sizes[i], linout = TRUE)})

my question is, i only know how to do it manually one by one, it means i have to run 100 times. Is there a way to run it once

Upvotes: 1

Views: 93

Answers (1)

Rich Scriven
Rich Scriven

Reputation: 99351

You could use replicate() and only create a single vector on each replication, instead of looping over a matrix. This just loops the expression block (code between the {} braces), evaluating it 100 times.

replicate(100, { 
    nn.sizes <- rbinom(6, 6, 0.5)

    <the rest of your code here> 

})

Upvotes: 2

Related Questions