Error: Cannot convert object of class ggsurvplotggsurvlist into a grob ,when i want to ggarrange a box plot on the left and ggsurvplot on the right

[ZihaoXingUP]

ggarrange(diff.exp.plot, ggsurvline.plot) Error: Cannot convert object of class ggsurvplotggsurvlist into a grob ,when i want to ggarrange a box plot on the left and ggsurvplot on the right

[jalavery]

Experiencing a similar issue trying to use par(mfrow=c(1,2))

library(survival)
library(survminer)

# example data frame
xx = data.frame(stim = c(2, 1:4),
                sts = rep(1, 5))

# survfit
oo = survfit(Surv(time = stim, event = sts) ~ 1, data = xx)

par(mfrow=c(1,2))

# plot 1
plot(oo)

# ggsurvplot
ggsurvplot(fit = oo, xlim=c(-2, 5))

The ggsurvplot is produced as an separate figure without the original plot on the side.

[ZihaoXingUP]

Experiencing a similar issue trying to use par(mfrow=c(1,2))

library(survival)
library(survminer)

# example data frame
xx = data.frame(stim = c(2, 1:4),
                sts = rep(1, 5))

# survfit
oo = survfit(Surv(time = stim, event = sts) ~ 1, data = xx)

par(mfrow=c(1,2))

# plot 1
plot(oo)

# ggsurvplot
ggsurvplot(fit = oo, xlim=c(-2, 5))

The ggsurvplot is produced as an separate figure without the original plot on the side.

i try your codes ,but it din't work

library(survival)
library(survminer)

# example data frame
xx = data.frame(stim = c(2, 1:4),
                sts = rep(1, 5))

# survfit
oo = survfit(Surv(time = stim, event = sts) ~ 1, data = xx)

par(mfrow=c(1,2))

# plot 1
plot(oo)

# ggsurvplot
ggsurvplot(fit = oo, xlim=c(-2, 5))

thanks your advice. but ggsurvplot based on ggplot , not on the par plot system. so ,it can't draw multi plots on a picture when i try your codes.

if i don't use ggsurvplot, i do can draw multi plots on one figure.

library(survival)
library(survminer)

# example data frame
xx = data.frame(stim = c(2, 1:4),
                sts = rep(1, 5))

par(mfrow=c(1,2))
oo = survfit(Surv(time = stim, event = sts) ~ 1, data = xx)

# plot a simple boxplot
plot(1:5,2:10)

#plot a survline plot
plot(oo)

[fufo64]

This is the solution: plot1<–datos %>% ggsurvplot(fit1,data=.) plot1<–datos %>% ggsurvplot(fit2,data=.)

ggarrange(plotlist=list(plot1$plot,plot2$plot), labels = c("A", "B",), ncol = 1, nrow = 2)

[jijitoutou]

This is the solution: plot1<–datos %>% ggsurvplot(fit1,data=.) plot1<–datos %>% ggsurvplot(fit2,data=.)

ggarrange(plotlist=list(plot1$plot,plot2$plot), labels = c("A", "B",), ncol = 1, nrow = 2)

Could you explain the code “plot1<–datos %>% ”？ This does not work in my computer. Thank you.

[jalavery]

This is the solution: plot1<–datos %>% ggsurvplot(fit1,data=.) plot1<–datos %>% ggsurvplot(fit2,data=.) ggarrange(plotlist=list(plot1$plot,plot2$plot), labels = c("A", "B",), ncol = 1, nrow = 2)

Could you explain the code “plot1<–datos %>% ”？ This does not work in my computer. Thank you.

plot1<–datos %>% ggsurvplot(fit1,data=.) is equivalent to: plot1<– ggsurvplot(fit1, data=datos) where datos is a data frame with survival information. Does that help to clarify?

[JennyCNS]

helps clarifying but it still doesn't solve the issue in my case.

Many thanks,

Jenny

[marine-ecologist]

Late to the answer here, but the output from ggsurv contains the combine output including plot and data, which is why it throws an error with ggarrange.

The answer by @fufo64 is correct in that specifying the $plot from the ggsurvplot output in ggarrange works, but there's no need to add them as a list or include the pipe in the ggsurv call.

Reproducible example here:

require("survival")
require(ggpubr)

fit<- survfit(Surv(time, status) ~ sex, data = lung)

a <- ggsurvplot(fit, data = lung)
b <- ggsurvplot(fit, data = lung, surv.median.line = "hv", palette = c("#E7B800", "#2E9FDF"))

ggarrange(a$plot,b$plot,ncol=2, widths=c(1,1), labels = c("a", "b"))

[hsiaoyi0504]

If you are using cowplot, plot_grid(a$plot,b$plot) also works