-
Notifications
You must be signed in to change notification settings - Fork 47
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
Add "purity" filter #19
Comments
I suggest instead operating directly on res,
and we can just extend susie_in_CS to compute purity (if X or LD is
provided) and
to filter if purity_threshold is provided.
so `susie_in_CS(res, LD_mat, purity_threshold)`
maybe `susie_get_CS` is better? verb (get) better than adjective (in)?
we could have a separate function purity(sets,LDmat) that does the work?
…On Fri, Jun 8, 2018 at 12:18 PM, gaow ***@***.***> wrote:
Currently I compute "purity" of sets separately based on LD and apply that
to susie sets (too bad I coded it in Python). We should implement it as a
separate function here. What should it be like? eg,
susie_get_sets(susie_in_CS(res), LD_mat = NULL, threshold = 0.2)
Or,
susie_get_sets(susie_in_CS(res), X = NULL, threshold = 0.2)
and we compute LD mat?
When LD mat is NULL, we just get the position of the variables for each
set and report them as an R list of sets. With LD_mat filter we
additionally compute minimal pairwise LD and remove the sets that fails the
threshold?
—
You are receiving this because you are subscribed to this thread.
Reply to this email directly, view it on GitHub
<#19>, or mute the thread
<https://github.com/notifications/unsubscribe-auth/ABt4xTYY3PW1u4qMLwEzM-8igxJ34j7Hks5t6rHbgaJpZM4UgqtC>
.
|
@stephens999 sorry I just cleaned up mostly my todo list on simulation so getting back to this one now. I think currently we use How about we then make BTW
I agree it is cleaner to have a separate function for developers when we want to explore how "purity" works. But I think now that we have some idea about it through separate studies, then here from user's prospective having less functions is better. So I'm voting for:
Does it sound Okay? |
I think correlation, X'X, is more natural to provide than squared correlation? |
So maybe Xcor ? |
Cool, |
Currently I compute "purity" of sets separately based on LD and apply that to susie sets (too bad I coded it in Python). We should implement it as a separate function here. What should it be like? eg,
Or,
and we compute LD mat?
When LD mat is NULL, we just get the position of the variables for each set and report them as an R list of sets. With LD_mat filter we additionally compute minimal pairwise LD and remove the sets that fails the threshold?
The text was updated successfully, but these errors were encountered: