Variance instead of SE
morrowcj opened this issue · 5 comments
morrowcj commented
In fitAR
, the returned values of SE
are, in fact, variances.
- The code should be something like
SE = diag(s2beta)^0.5
- check if other versions of fitX have this problem.
Rapsodia86 commented
Hi,
any plans to have it updated?
arives commented
Monika,
Yep, Clay is just about to. Everything else, including the t-scores and p-values, is fine; it’s really just a matter of taking the square-root of the printed SEs for the fitAR and fitCLS functions.
Sorry about this, Tony
…__________________________
Anthony R. Ives (he/him/his)
UW-Madison
459 Birge Hall
608-262-1519
From: Monika Anna Tomaszewska ***@***.***>
Date: Monday, November 14, 2022 at 3:42 PM
To: morrowcj/remotePARTS ***@***.***>
Cc: Subscribed ***@***.***>
Subject: Re: [morrowcj/remotePARTS] Variance instead of SE (Issue #2)
Hi,
any plans to have it updated?
—
Reply to this email directly, view it on GitHub<#2 (comment)>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/ACYX6LAE4J4JXLRBNKH3SJDWIKW2RANCNFSM576MAQYA>.
You are receiving this because you are subscribed to this thread.Message ID: ***@***.***>
Rapsodia86 commented
Hello,
Thanks for the quick answer!
Yes, I have explored the code to find where to apply the correction (or to be aware about the output).
But I appreciate the fact you are going to do this! Thanks!
morrowcj commented
@Rapsodia86 and @arives,
I just wanted to alert you that I've corrected this with PR #4 which has been merged into the live branch, as soon as the build checks finished running. Sorry it took so long to get to this simple fix!
Having addressed the problem, I'm closing this issue now.
Rapsodia86 commented
Thank you for the update and fixing the issue!