[R-meta] Error message in ML meta-analysis

[Viechtbauer, Wolfgang (SP)]

Dear G.J.,

Based on what you show, I cannot diagnose the problem.

What does class(microtrialmerge) show? How about class(microtrialmerge$smd_var)?

Even better, can you create a small fully reproducible example that creates this error?

Best,
Wolfgang

-----Original Message-----
From: R-sig-meta-analysis [mailto:r-sig-meta-analysis-bounces using r-project.org] On Behalf Of G.J. Melendez-Torres
Sent: Wednesday, 06 June, 2018 1:47
To: r-sig-meta-analysis using r-project.org
Subject: [R-meta] Error message in ML meta-analysis

Hi everyone-after going back and running a multilevel meta-analysis that worked a few months ago, I got this error message:

> timeout.obs<-rma.mv(smd_rev, smd_var, random= ~esid | code, data=microtrialmerge)
Error in if (is.vector(V) || nrow(V) == 1L || ncol(V) == 1L) V <- diag(as.vector(V),  :
  missing value where TRUE/FALSE needed

This is rather vexing because a) these exact same data and same command have run before, and b) I have checked and rechecked and cannot find any reason why this might be the case.  The yi and vi vectors are both k x 1 with no missing values, and I can't seem to find a logical in the metafor guidance that would require specification for this basic meta-analysis

Any assistance would be gratefully received...

G.J.

G.J. Melendez-Torres DPhil RN MFPH
Senior Lecturer | Uwch-ddarlithydd

DECIPHer, School of Social Sciences | DECIPHer, Ysgol Gwyddorau Cymdeithasol
Cardiff University | Prifysgol Caerdydd
1-3 Museum Place | 1-3 Plas yr Amgueddfa
Cardiff | Caerdydd
CF10 3BD

Email | Ebost: Melendez-TorresG using cardiff.ac.uk<mailto:Melendez-TorresG using cardiff.ac.uk>
Tel. | Ff�n.: +44 (0) 29208 79106
Follow DECIPHer on Twitter | Dilynwch DECIPHer ar Twitter: @DECIPHerCentre<https://twitter.com/DECIPHerCentre>
Follow me on Twitter | Dilynwch fi ar Twitter: @gjmelendez