Switch between mgcv and gam

[scottkosty]

A common desire is to switch between using the mgcv and gam packages, as noted here for example:
http://stackoverflow.com/questions/22126611/r-package-conflict-between-gam-and-mgcv

As noted in ?detach,

Further, registered S3 methods from the namespace will not be removed.

However, perhaps switchr knows some trick to get around this.

[scottkosty]

@gmbecker I just wanted to check in to see if I should maintain a sliver of hope that we could find a solution to this issue within the context of switchr. No problem at all if you are too busy to take another look at it. I'm just curious if you happen to have any updated thoughts on if it is even possible.

By the way, I'm glad to see that the development of this package is still going strong!

[gmbecker]

I will look at this again when I get a minute, but unfortunately, from what I recall, the fact that S3 methods are not unregistered when packages are unloaded, and will actually re-load the packages they are from when hit appears to be intentional in the current design of how how S3 works. So long as this is the case, I think that any solution to your particular problem by any piece of code, in switchr or elsewhere, is impossible. @lawremi is this your impression as well? The issue is that two packages both define S3 methods for the same generic, so you can't "switch between them" by unloading one and then loading the other.

…

On Sun, Mar 19, 2017 at 10:04 AM, Scott Kostyshak ***@***.***> wrote: @gmbecker <https://github.com/gmbecker> I just wanted to check in to see if I should maintain a sliver of hope that we could find a solution to this issue within the context of switchr. No problem at all if you are too busy to take another look at it. I'm just curious if you happen to have any updated thoughts on if it is even possible. By the way, I'm glad to see that the development of this package is still going strong! — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#2 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AA3dsaZhWeqOEOoFwVBLW7UPGYdqah1Pks5rnWAKgaJpZM4Diwp1> .

-- Gabriel Becker, PhD Associate Scientist (Bioinformatics) Genentech Research

[scottkosty]

@gmbecker thanks for taking another look and for your thoughts. No problem if there is nothing we can do. I cannot think of a workaround but I'll keep thinking.