[BioC] limma normalization warning

Gordon Smyth smyth at wehi.edu.au
Mon Sep 26 03:02:22 CEST 2005


>Date: Tue, 20 Sep 2005 10:55:19 -0500
>From: <yjzhang at uiuc.edu>
>Subject: [BioC] limma normalization warning
>To: bioconductor at stat.math.ethz.ch
>
>I am using limma normalizeWithinArrays with method
>="robustspline" and other default settings. I got the
>following warning: rlm failed to converge in 20 steps in:
>rlm.default(x, y, weights = w, method = method).

No problem. Ignore it.

>I googled and found someone solved the problem by increasing
>the parameter "maxit". When I tried to pass "maxit=21", I
>got an error message: "unused argument(s) (maxit ...)".

Pretty self-explanatory I would have thought.

Every function has a help page which explains exactly what the arguments 
are and what they mean. You are expected to read it before you ask a 
question on the list. No exceptions.

Gordon

>YJ Zhang



More information about the Bioconductor mailing list