extract data.frame name and dependent variable name from nls model in R

Question

Say I have I fit an nls model to some data, d , in R.

#fake data
x<- seq(1,10,by=0.1)
y<- 1*(x / (2 + x))
y<- y + rnorm(0, 0.05, n =length(y))
d<- data.frame(x,y)

#fit the nls model
model <- nls(y ~ v1*(x/(v2 + x)), start=c(v1=0.9,v2=2.1))

I want to return from the model object the name of the data frame ( d ), and the name of the dependent varaible ( y ).

Answer 1

You can get these values from the model object as follows:

# Data frame name
as.character(model$data)

# Dependent variable name
as.character(formula(model)[2])

How did I know that, you ask? Well, I didn't, but you can often figure out things like this by looking at the structure of the model object (type str(model) in the console) returned by an R function.

For example, in this case, model is a list with six elements. The first element, called m is itself a list. Here are the first few lines of output:

str(model)

 List of 6 $ m :List of 16 ..$ resid :function () ..$ fitted :function () ..$ formula :function () 

Note that one of the elements of m is called formula and it's a function. So, I tried typing formula(model) in the console and got this:

 y ~ v1 * (x/(v2 + x)) 

Then I typed str(formula(model)) to see what kind of object formula() was actually returning:

 Class 'formula' length 3 y ~ v1 * (x/(v2 + x)) ..- attr(*, ".Environment")=<environment: R_GlobalEnv> 

Notice that it returns an object with three elements ( length 3 ). The second element happens to contain the dependent variable name. Wrapping it in as.character changes it into a string, which is what I assume you want.

Now for the data frame. Note that one of the list elements returned by str(model) is called data . So I typed model$data in the console and got this:

 d 

Once again, I wrapped it in as.character to turn the output into a string.

There may be better ways to approach this, but don't underestimate what you can accomplish just by poking around inside the object returned by a function.