CODEWITHSUNDEEP
rggplot2
KT_1
KT_1

Reputation: 8494

Using melt in ggplots2

For a sample dataframe:

df <-structure(list(Type = structure(c(1L, 1L, 1L, 1L, 1L, 1L, 1L, 
1L, 2L, 2L, 3L, 3L, 3L, 3L), .Label = c("A", "B", "C"), class = "factor"), 
    RSL = c(-1.84, 7.68, 18.4, 13, 39.8, 98.4, 129.9, 94.9, 138, 
    142, 10, 50, 60, 80), Age = c(1181.5, 4633, 5075.5, 5741.5, 
    8430.5, 9784, 10095, 10366, 10095, 10095, 6000, 13500, 13000, 
    12000)), .Names = c("Type", "RSL", "Age"), class = "data.frame", row.names = c(NA, 
-14L))

I want to produce a graph with Age plotted against RSL (this is a single graph which forms part of many faceted graphs).

library(ggplot2)
library("reshape2")
g <- ggplot (df, aes(x=Age, y=RSL, shape = Type, colour=Type)) +
  geom_point(size=1.5) +
  scale_shape_manual(values=c(19,19,19,19,19))+ # sets shape of points
  scale_colour_manual(values=c("red", "black", "blue")) #sets colour of points
g

But instead of A, B and C all being points, I want C to be a straight line (no points).

How is the best way to deal with this? I believe I need to melt my data:

df_long <- melt(df, id="Type")

... but can't work out how to assemble my graph?

Upvotes: 1

Views: 85

Answers (2)

pogibas
pogibas

Reputation: 28379

Pass data with C to a different geom layer (in your case geom_line) with data = subset(df, Type == "C".

library(ggplot2)
ggplot(subset(df, Type != "C"), aes(Age, RSL)) +
    geom_point(aes(shape = Type, colour = Type), size = 1.5) +
    geom_line(data = subset(df, Type == "C"), color = "blue") +
    scale_shape_manual(values=c(19,19,19,19,19))+ 
    scale_colour_manual(values = c("red", "black"))

enter image description here

Upvotes: 2

missuse
missuse

Reputation: 19756

This should also work and the legend will display all 3 too:

ggplot () +
      geom_point(data = df, aes(x=Age, y=RSL,  colour=Type), size=0, alpha = 0)+
      geom_point(data = df[df$Type!="C",], aes(x=Age, y=RSL, colour = Type), size=1.5) +
      geom_line(data = df[df$Type=="C",], aes(x=Age, y=RSL), size=0.7, color="blue")+
      scale_colour_manual(values=c("red", "black", "blue"))

enter image description here

Upvotes: 3

Related Questions