Combining multiple data frames in one plot using ggplot2

Question

I'd like to plot multiple data frames on one plot using ggplot2 . I have about 30 data frames, and each data frame contains the same variables (x=value, y=index, and standard error) like the example below.

eg)

df1 <-
       value        SE              index
1  -3.447418 0.4308221           19.00978
2  -3.999097 0.4308221          147.79562
3  -4.316288 0.4308221          268.78998
4  -4.449099 0.4308221          332.87519
5  -3.696987 0.4308221          447.74797
6  -3.313633 0.4308221          565.46903
7  -3.335039 0.4308221          709.58848
8  -2.486115 0.4308221          838.49382
9  -1.230000 0.4308221          993.37466
10 -2.558116 0.4308221         1150.04461

df2
.
.
.
.
.
df30

I want to plot all 30 data frames with different colors with error bars shown for each value. I am new to R and I was barely able to plot a single data frame with an error bar with the following code.

A plot image

df1 = read.table("data_1.txt", header=TRUE, sep="\t")

p = ggplot(df1, aes(x=index, y=value)) + 
    geom_line(size=1, colour = "coral") + 
    geom_point(size=2.5, colour = "coral") + 
    ylab("value") + xlab("index")

gp = p + scale_y_continuous(limits=c(-5.5, 0.5), breaks=seq(-5.5, 0.5, 0.5)) + 
     scale_x_continuous(limits=c(0, 1530), breaks=seq(0, 1530, 250)) + 
     geom_errorbar(aes(ymax=value+sd, ymin=value-sd), width=20, size=0.2, colour = "black") + 
     theme_classic()

plot(gp)

I have been looking into Displaying multiple data frames in ggplot2 and R : ggplot2 plot several data frames in one plot , but I could not figure out the best solution. Can someone tell me how I would be able to do what I'm looking for?

Thank you very much for your help in advance.

Answer 1

If you create a list of the dataframes you will be able to use bind_rows from the dplyr package, like

    df_plot <- bind_rows(df_list, .id = original_df)

    ggplot(df_plot, aes(index, value, group = original_df, colour = original_df) +
     geom_line(size=1) + 
     geom_point(size=2.5) + 
     labs(x = 'index', y = "value") + 
     scale_y_continuous(limits=c(-5.5, 0.5), breaks=seq(-5.5, 0.5, 0.5)) + 
     scale_x_continuous(limits=c(0, 1530), breaks=seq(0, 1530, 250)) + 
     geom_errorbar(aes(ymax=value+sd, 
                       ymin=value-sd), 
                   width=20,
                   size=0.2, 
                   colour = "black") + 
 theme_classic()

Answer 2

You may add to each dataframe a new column identifying it.

df1$id <- "df1"  # or any other description you want
df2$id <- "df2"

Then you join all dataframes:

df.all <- rbind(df1, df2, ..., df30)

Now you can plot all the data together by using the group= and color= aesthetics options:

p = ggplot(df.all, aes(x=index, y=value, group=id, color=id)) + 
    geom_line(size=1, colour = "coral") + 
    geom_point(size=2.5, colour = "coral") + 
    ylab("value") + xlab("index") +
    scale_y_continuous(limits=c(-5.5, 0.5), breaks=seq(-5.5, 0.5, 0.5)) + 
    scale_x_continuous(limits=c(0, 1530), breaks=seq(0, 1530, 250)) + 
    geom_errorbar(aes(ymax=value+sd, ymin=value-sd), width=20, size=0.2, colour = "black") + 
    theme_classic()

I am afraid, though, that 30 overlapping data sets on a single plot may look crowded. You may also want to try with facets, which are small panels for each individual plot, as indicated below.

p = ggplot(df.all, aes(x=index, y=value)) + 
    geom_line(size=1, colour = "coral") + 
    geom_point(size=2.5, colour = "coral") + 
    ylab("value") + xlab("index") +
    scale_y_continuous(limits=c(-5.5, 0.5), breaks=seq(-5.5, 0.5, 0.5)) + 
    scale_x_continuous(limits=c(0, 1530), breaks=seq(0, 1530, 250)) + 
    geom_errorbar(aes(ymax=value+sd, ymin=value-sd), width=20, size=0.2, colour = "black") + 
    theme_classic() +
    facet_wrap(~id)