CODEWITHSUNDEEP
rggplot2
Aru
Aru

Reputation: 33

How can I plot this complex ggplot with multiple layers together?

I am trying to figure out how to create this complex ggplot using the dataset below:

type a b d points end start
A -0.32 -0.02 0 1.08 1.14 1.03
A -0.32 -0.02 2 0.39 0.45 0.32
A -0.32 -0.02 4 0.25 0.32 0.17
A -0.32 -0.02 6 0.06 0.07 0.04
B -0.19 -0.02 0 1.07 1.12 1.01
B -0.19 -0.02 2 0.55 0.58 0.52
B -0.19 -0.02 4 0.43 0.48 0.39
B -0.19 -0.02 6 0.17 0.20 0.15

The plot will be like this:

enter image description here

For each "type", there will be a curve, a set of 4 points, and a set of 4 segments.

  1. Each curve will be a curve of the the function [exp(ax + bx^2)], where a and b are in the dataset. I think it can be achieved by:
geom_function(fun = function(x) exp(a*x + b*x^2))
  1. The points can be plotted from the variable "points" like this:
geom_point(aes(x = d, y = points))
  1. The segments can be plotted from the start and end variables like this:
geom_segment(aes(x = d, y = start, xend = d, yend = end))
  1. The y axis is in a log10 scale, which can be achieved by adding:
scale_y_log10()

I think I know the individual details, but unfortunately I am not being able to plot it after I put everything together. The curves of the functions are not coming up, and I am getting errors. Can anyone please help?

Thank you so much!!!!

Upvotes: 1

Views: 153

Answers (1)

CzechInk
CzechInk

Reputation: 473

Here is an alternative approach that uses geom_line() after generating the y data:

library(dplyr)
library(ggplot2)

# replicating your data
df <- data.frame(
  type = rep(c("A","B"), each = 4),
  a = rep(c(-.32, -0.19), each = 4),
  b = rep(-.02, 8),
  d = rep(c(0:3)*2, 2),
  points = c(1.08, 0.39, 0.25, 0.06, 1.07, 0.55, 0.43, 0.17),
  end = c(1.14, 0.45, 0.32, 0.07, 1.12, 0.58, 0.48, 0.20),
  start = c(1.03, 0.32, 0.17, 0.04, 1.01, 0.52, 0.39, 0.15))

# new data.frame for plotting the lines
n_seq <- 100 # how many y-values to generate
df_sim <- data.frame(d = rep(with(df, seq(min(d), max(d), length.out = n_seq)), 2),
                     type = rep(c("A", "B"), each = n_seq),
                     a = rep(c(-.32, -0.19), each = n_seq),
                     b = rep(-.02, n_seq*2)) %>%
  mutate(y_hat = exp(a*d + b*d^2))

ggplot(data = df, aes(x = d, y = points, color = type)) +
  geom_pointrange(aes(ymin = end, ymax = start)) + # error bar & points in one step
  geom_line(data = df_sim, aes(y = y_hat, x = d, color = type)) + # instead of geom_function
  scale_y_log10() +
  ylab("y") # so that y-axis is labeled like the ex

Upvotes: 1

Related Questions