I draw a forest plot with non-default annosym values:
res <- rma(ai=tpos, bi=tneg, ci=cpos, di=cneg, data=dat.bcg, measure = 'OR', method='DL',
slab=paste(author, year))
forest(res,
annosym=c(' (','; ',')'),
ylim=c(-2,16)
)
And try to add some sub-analysis results with the same annosym specification:
dat.bcg.rand <- dat.bcg[dat.bcg$alloc=='random',]
res.rand <- rma.uni(ai=tpos, bi=tneg, ci=cpos, di=cneg, data=dat.bcg.rand, measure = 'OR', method='DL',
slab=paste(author, year))
addpoly(res.rand, mlab='Random alloc.',atransf=exp, annosym=c(' (','; ',')'))
to see a warning:
Warning messages:
1: In text.default(...) : "annosym" is not a graphical parameter
2: In segments(...) : "annosym" is not a graphical parameter
3: In segments(...) : "annosym" is not a graphical parameter
4: In segments(...) : "annosym" is not a graphical parameter
5: In polygon(...) : "annosym" is not a graphical parameter
6: In text.default(...) : "annosym" is not a graphical parameter
and a plot that mixes two annosym specifications:
Do you have any workarounds for this? I really, really need to change default annosyms in addpoly().
