[R-meta] question on forestplot by Metafor

[Viechtbauer, Wolfgang (NP)]

Dear Frank,

See

https://wviechtb.github.io/metafor/reference/forest.rma.html

and in particular the 'addpred' argument.

Best,
Wolfgang

>-----Original Message-----
>From: R-sig-meta-analysis [mailto:r-sig-meta-analysis-bounces using r-project.org] On
>Behalf Of Frank van Boven via R-sig-meta-analysis
>Sent: Friday, 28 April, 2023 19:25
>To: R Special Interest Group for Meta-Analysis
>Cc: Frank van Boven
>Subject: [R-meta] question on forestplot by Metafor
>
>Dear all,
>
>May be a little bit a not expected question, I would like to produce a forestplot
>without the polygon (diamond) at the bottom.
>The rationale for not presenting an overall-effect follows from non-statistical
>viewpoints.
>By erasing it, the figure loses quality (see figure).
>Is there a way to chose transparant or white for plotting the polygon (without
>losing the dashed ref-line).
>Enclosed part of the codes.
>
>par(mar=c(4,4,1,2))
>forest(res, slab = paste(dat44$Author, dat44$Year, sep = ", "), xlim = c(-20,
>10), at=c(-3.0, -1.0, 0, 2.0), refline = 0,
>rows=c(25:22, 18:16, 12:12, 9:5, 1:1),ylim=c(-1, 30), cex = 0.75,
>ilab=cbind(dat44$n1, dat44$n2), ilab.xpos= c(-9, -7))
>
>op <- par(cex = 0.75, font = 2)
>text(-20, 30, "Asthma symptom scores", pos = 4)
>text(-20, 29, "Author; Year", pos = 4)
>text (-10.5, 27, "Sample sizes",pos = 4)
>text (c(-10, -8), 26, c("n1","n2"), font = 1, pos = 4)
>text(10, 29, "SMD [95% CI]", pos = 2)
>text(-20, c(26, 19, 13, 10,2), pos=4, c("Concurrent - One barrier", "Concurrent -
>Two barriers",
>"Concurrent - Three barriers", "Concurrent - Other versions", "Air purification -
>Nocturnal"))
>text(-20,-1, "RE Model", pos  = 4, cex = 1.0, font = 1, col = "white")
>
>res.1 <- rma(yi, vi, subset=(Strategy=="Concurrent_one"), data = dat44)
>res.1
>
>addpoly(res.1, cex = 1.0, font = 2, row = 21.0)
>text(-20,21, "RE Model", pos  = 4, cex = 1.0, font = 2, col = "white")
>text(-20,21, "     SMD for subgroup", pos  = 4, cex = 1.0, font = 2)
>
>par(op)
>
>Looking forward for your reply.
>
>Kind regards,
>
>Frank van Boven