Error in terms.formula(formula) : '.' in formula and no 'data' argument

Question

I'm tring to use neuralnet for prediction.

Create some X:

x <- cbind(seq(1, 50, 1), seq(51, 100, 1))

Create Y:

y <- x[,1]*x[,2]

Give them a names

colnames(x) <- c('x1', 'x2')
names(y) <- 'y'

Make data.frame:

dt <- data.frame(x, y)

And now, I got error

model <- neuralnet(y~., dt, hidden=10, threshold=0.01)

error in terms.formula(formula) : '.' in formula and no 'data' argument

For example, in lm(linear model) this is worked.

`neuralnet` is doing a lot os messing with the formula via non-exported function `neuralnet:::generate.initial.variables`. There is a bug in that function. I suggest you contact the maintainer and send them this example or a link to the question. — Gavin Simpson, Jul 22 '13 at 18:28

score 50 · Accepted Answer · answered Jul 22 '13 at 18:33

50

As my comment states, this looks like a bug in the non-exported function neuralnet:::generate.initial.variables. As a work around, just build a long formula from the names of dt, excluding y, e.g.

n <- names(dt)
f <- as.formula(paste("y ~", paste(n[!n %in% "y"], collapse = " + ")))
f

## gives
> f
y ~ x1 + x2

## fit model using `f`
model <- neuralnet(f, data = dt, hidden=10, threshold=0.01)

> model
Call: neuralnet(formula = f, data = dt, hidden = 10, threshold = 0.01)

1 repetition was calculated.

        Error Reached Threshold Steps
1 53975276.25     0.00857558698  1967

answered Jul 22 '13 at 18:33

Gavin Simpson

170,508
25
396
453

sorry, but I can't understand. Why error value is so big? So I did, but got constant predicted value. Can you help me? – luckyi Jul 23 '13 at 17:37
@luckyi That is likely a statistics question and is not suited to [so] Try asking on [stats.se]. – Gavin Simpson Jul 23 '13 at 17:39
1

This also happens in package RMS for ols(). The same fix works there too – Chris Dec 12 '14 at 22:49
1

@GavinSimpson - Thanks for the solution, Gavin! I've encountered the same formula error in the new mpath package. – RobertF Jun 23 '15 at 14:19
it doesn't remove the y values from the left hand side of formula. It comes in the end. – Arpit Sisodia May 14 '18 at 08:00

score 13 · Answer 2 · answered Jun 29 '17 at 13:04

13

Offering a simpler alternative to the previous answer, you can create a formula from names of dt using reformulate():

f <- reformulate(setdiff(colnames(dt), "y"), response="y")

reformulate() doesn't require the use of paste() and automatically adds the terms together.

answered Jun 29 '17 at 13:04

Megatron

15,909
12
89
97

I've been searching for this,,,,, – kurtkim Sep 28 '22 at 13:38

Gorka · Answer 3 · 2021-09-21T15:08:29.610

-1

To expand a formula

f <- formula(terms(f, data= dt))

or even shorter

f <- formula(dt, f)

where f is the formula and dt is the data.

For instance, the original formula could be:

f <- as.formula("y ~ .")

edited Sep 21 '21 at 15:08

answered Sep 21 '21 at 12:32

Gorka

1,971
1
13
28

Really confusing answer – Julien Oct 16 '22 at 07:10

Error in terms.formula(formula) : '.' in formula and no 'data' argument

3 Answers3

Linked