I'm trying to conduct a network meta-analysis using the netmeta package, but keep getting this error message. All the standard errors and treatment effects are entered and non-zero. Does anyone know what causes this, and if so what solutions exist?
> netmeta(TE=est, seTE=se, treat1=arm1l, treat2=arm2l, data=ldlc_1012_2, studlab=study)
Error in eigen(S %*% P %*% S, symmetric = TRUE, only.values = TRUE) :
infinite or missing values in 'x'
Thanks!