Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Feature selection - parameter naming conflict #1066

Closed
srp33 opened this issue Aug 5, 2016 · 12 comments
Closed

Feature selection - parameter naming conflict #1066

srp33 opened this issue Aug 5, 2016 · 12 comments
Assignees
Labels

Comments

@srp33
Copy link

srp33 commented Aug 5, 2016

I really enjoy working with mlr, but I've run into a little problem, and I'm not sure how to get around it. I'm hoping you can help. I'm trying to do feature selection with the rf.min.depth function. I want to specify a value for the "method" parameter within the underlying library ( randomForestSRC::var.select). But there is also a "method" parameter in the generateFilterValuesData function for the learner name. Below is some example code along with the error I'm getting. I'm running R-3.3.1 and mlr version 2.9.

library(mlr)

parameterValues <- c(method="vh", conservative="low")
task <- makeClassifTask(data = iris, target = "Species")
fv <- generateFilterValuesData(task, method = "rf.min.depth", more.args=list("rf.min.depth"=parameterValues))

Error in randomForestSRC::var.select(getTaskFormula(task), getTaskData(task), :
formal argument "method" matched by multiple actual arguments

Any suggestions?

@jakobbossek
Copy link
Contributor

The problem is that in an internal function method is set directly and via .... Definitely a bug. Going to fix this soon.

@berndbischl
Copy link
Sponsor Member

@jakobbossek
please close this when it is unit tested

@jakobbossek
Copy link
Contributor

Added test in b8c46bc. Closing.

@srp33
Copy link
Author

srp33 commented Aug 11, 2016

Sorry for the dumb question, but what can I do if I want to apply this patch to my own analysis? I am currently using version 2.9.

@larskotthoff
Copy link
Sponsor Member

You can install the development version of mlr with devtools:

devtools::install_github("mlr-org/mlr")

@srp33
Copy link
Author

srp33 commented Aug 11, 2016

Will this be incorporated into version 2.10? If so, how soon do you expect that to be released? Thanks again!

@larskotthoff
Copy link
Sponsor Member

Yes and we expect a release very soon.

@berndbischl
Copy link
Sponsor Member

monday my holiday starts. i want to push to CRAN before this.

they should hopefully allow a 2nd release in 10 days.

@srp33
Copy link
Author

srp33 commented Aug 11, 2016

What would be included in the 2nd release?

@larskotthoff
Copy link
Sponsor Member

We are still working on this.

@berndbischl
Copy link
Sponsor Member

What would be included in the 2nd release?

simply read the NEWS in the GH repo. you can see what we are changing there.

@srp33
Copy link
Author

srp33 commented Aug 29, 2016

Sorry to keep this thread going, but do you know how soon version 2.10 will be live on CRAN? It is still showing as 2.9. Thanks!

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
Projects
None yet
Development

No branches or pull requests

4 participants