Reputation:
160
Status:
Trusted
Location:
Heidelberg
Last seen:
3 weeks, 6 days ago
Joined:
3 years, 6 months ago
Email:
n*****************@gmail.com

<prev • 8 results • page 1 of 1 • next >
1
59
views
1
... That is an excellent question! The weights are indeed calibrated towards a specific alpha value. So if you run IHW with different alphas, you will get different weights in general. Furthermore, if you run IHW at a specific alpha, but then use another alpha' cutoff on the adjusted p-values this wil ...
written 4 weeks ago by Nikos Ignatiadis160
2
228
views
2
... Without any further information about your data (e.g. maybe discrete tests), the results you obtain from qvalue appear plausible to me. What you are discovering here is that in many cases, from a statistical perspective, the problem of detection is much easier than the problem of localization. In ot ...
written 9 months ago by Nikos Ignatiadis160
2
258
views
2
Answer: A: Libgomp error in IHW
... Hello! Indeed this appears to be related to the parallelization in Lpsymphony. Can you give some more context about your problem? How many p-values do you have and what are the covariates? Furthermore, as a basic sanity check, can you run the IHW examples (either in the documentation or the vignette ...
written 12 months ago by Nikos Ignatiadis160
1
275
views
1
... Hi Paul, it's a good question. As is so often the case, this is just a bias/variance/computation time tradeoff. Here by bias-variance I refer to the estimation of the weight function from the rest of the folds and applying it to the held-out fold. In general, in the default choices in the IHW packa ...
written 21 months ago by Nikos Ignatiadis160
1
505
views
1
Comment: C: IHW Package for R 3.2.5
... Mike's suggestion is the most elegant one and I'd recommend you go that route. If you really can't move to another version, you could probably manually download IHW and its dependencies and manually change the R version requirements in the DESCRIPTION file. It should probably work (although not te ...
written 23 months ago by Nikos Ignatiadis160
1
375
views
1
... Hi Mike, I am afraid this will largely be a non-answer, since right now the support for multivariate covariates is minimal: You have to manually bin covariates into higher dimensional rectangles, with the obvious repercussions in terms of the curse of dimensionality. One thing you could do is just ...
written 2.3 years ago by Nikos Ignatiadis160
2
701
views
2
... A key diagnostic to figure out whether the cutoff of 60% is reasonable would be to check how many rejections you get after applying this procedure (and possibly also compare to the baseline without any filtering). If the number of rejections is very low, then I would worry. If the number of rejectio ...
written 2.9 years ago by Nikos Ignatiadis160
1
587
views
1
... Hi, this is the first IHW question on the support forums and a good one! Actually there are two distinct parts to your question: 1. Can the direction of the effect be used as a covariate? It depends on the application, but in many cases you can. In particular, it should be valid (i.e. conditional ...
written 2.9 years ago by Nikos Ignatiadis160

#### Latest awards to Nikos Ignatiadis

Scholar 21 months ago, created an answer that has been accepted. For A: genefilter filtered out 60% of the original genes?
Teacher 21 months ago, created an answer with at least 3 up-votes. For A: Possible to incorporate the sign or direction of an effect in the IHW covariate?
Scholar 2.9 years ago, created an answer that has been accepted. For A: Possible to incorporate the sign or direction of an effect in the IHW covariate?
Teacher 2.9 years ago, created an answer with at least 3 up-votes. For A: Possible to incorporate the sign or direction of an effect in the IHW covariate?
Scholar 2.9 years ago, created an answer that has been accepted. For A: Possible to incorporate the sign or direction of an effect in the IHW covariate?

Content
Help
Access

Use of this site constitutes acceptance of our User Agreement and Privacy Policy.