Legend in ggplot2

[juanriera]

Hello,
I'm trying to recreate this plot using ggplot2. I have been quite successful until now but I'm unable to put the legend like the example. How should I get the legend as in cited plot (I know how to do a legend in ggplot, I just want to get the same legend and position if possible.
This is my code:

# ----------------------------------------------------------------------------
# Mi versión con ggplot
library(tidyverse)
library(showtext)

## Loading Google fonts (https://fonts.google.com/)
font_add_google("Gochi Hand", "gochi")
font_add_google("Schoolbell", "bell")
font_add_google("Covered By Your Grace", "grace")
font_add_google("Rock Salt", "rock")

## Automatically use showtext to render text for future devices
showtext_auto()

## Tell showtext the resolution of the device,
## only needed for bitmap graphics. Default is 96
showtext_opts(dpi = 96)

set.seed(123)
x  <- rnorm(10)
y  <-  1 + x + rnorm(10, sd = 0.2)
y[1] <-  5
mod <- lm(y ~ x)

df <- data.frame(x,y)

df |> ggplot(aes( x = x, y = y)) +
  geom_point(size = 3, colour = "steelblue") + # colores: cadetblue4, dodgerblue3, firebrick4, 
  geom_smooth(method = 'lm', formula = y~x, se = F, colour = "black") +
  geom_abline(aes(slope = 1, intercept = 1), colour = "red") +
  labs(
    x = "x variable",
    y = "y variable",
    title = "Draw Plots Before You Fit A Regression",
  ) +
  geom_text(
    aes(x, y), 
    data = data.frame(x = -0.5, y = 4.75), 
    label = "This is the outlier",
    colour = "steelblue",
    family = "grace",
    size = 7
  ) + 
  geom_text(
    aes(x,y),
    data = data.frame(x = 1, y = 1.5), 
    label = expression(paste("True model: ", y == x + 1)),
    family = "rock",
    angle = 30,
    size = 6, 
    colour = "red" 
  ) +
  geom_text(
    aes(x,y),
    data = data.frame(x = 0, y = 2), 
    label = expression(paste("OLS: ", hat(y) == 0.79 * x + 1.49)),
    family = "rock",
    angle = 22,
    size = 7, 
    colour = "black" 
  ) +
  theme_bw() + 
  theme(
    plot.title = element_text(size = 24, family = "bell"),
    axis.text.x = element_text(size=14, family = "gochi"),
    axis.text.y = element_text(size=14, family = "gochi"),
    axis.title.x = element_text(size=24, family = "gochi"),
    axis.title.y = element_text(size=24, family = "gochi"),
  )

This is the plot rendered:

[technocrat]

This doesn't have what ggplot thinks of as a legend. The plot shows a "title" on top and "labels" for the x-axis and y-axis. Where is what you are looking to change?

[juanriera]

It's true, as it is, code doesn't have any legend element to plot. I am looking for some layer or geom to add to the code posted to get that legend. It is only an exercise; I am practising with ggplot just to learn. I have tried several different ones, but I haven't found any geom that adds the legend as it is on the original plot. R basic plot functions add that legend with just

legend("topright", legend = c("Truth", "OLS"), col = c("red", "black"), lty = 1)

This is the legend in original plot and that's what I want to get.

[technocrat]

We are still not on the same page for legend I think. Here's what one looks like

[juanriera]

Thanks for answering. Yeah, I know what a legend is, of course. I think I have been a bit clearer maybe on my last post

[technocrat]

Posing the question is always the hard part, ain't it? Still unclear on the what though that you want to add.

[juanriera]

Getting closer to it, only x axis ticks labels missing
The code:

# ----------------------------------------------------------------------------
# Mi versión con ggplot
library(tidyverse)
library(showtext)
library(scales)

## Loading Google fonts (https://fonts.google.com/)
font_add_google("Gochi Hand", "gochi")
font_add_google("Schoolbell", "bell")
font_add_google("Covered By Your Grace", "grace")
font_add_google("Rock Salt", "rock")

## Automatically use showtext to render text for future devices
showtext_auto()

## Tell showtext the resolution of the device,
## only needed for bitmap graphics. Default is 96
showtext_opts(dpi = 96)

set.seed(123)
x  <- rnorm(10)
y  <-  1 + x + rnorm(10, sd = 0.2)
y[1] <-  5
mod <- lm(y ~ x)

df <- data.frame(x,y)

df |> ggplot(aes( x = x, y = y)) +
  geom_point(size = 3, colour = "steelblue") + # colores: cadetblue4, dodgerblue3, firebrick4, 
  geom_smooth(aes(colour = "black"), method = 'lm', formula = y~x, se = F) +
  geom_abline(aes(slope = 1, intercept = 1, colour = "red")) +
  scale_colour_manual(values=c("black", "red"), labels = c("OLS", "Truth")) +
  scale_x_continuous(labels = label_number(accuracy = 0.1)) +
  labs(
    x = "x variable",
    y = "y variable",
    title = "Draw Plots Before You Fit A Regression",
  ) +
  geom_text(
    aes(x, y), 
    data = data.frame(x = -0.5, y = 4.7), 
    label = "This is the outlier",
    colour = "steelblue",
    family = "grace",
    size = 7
  ) + 
  geom_text(
    aes(x,y),
    data = data.frame(x = 1, y = 1.2), 
    label = expression(paste("True model: ", y == x + 1)),
    family = "rock",
    angle = 30,
    size = 6, 
    colour = "red" 
  ) +
  geom_text(
    aes(x,y),
    data = data.frame(x = 0, y = 2), 
    label = expression(paste("OLS: ", hat(y) == 0.79 * x + 1.49)),
    family = "rock",
    angle = 22,
    size = 7, 
    colour = "black" 
  ) +
  theme_bw() + 
  theme(
    plot.title = element_text(size = 22, family = "bell"),
    axis.text.x = element_text(size=18, family = "gochi"),
    axis.text.y = element_text(size=18, family = "gochi"),
    axis.title.x = element_text(size=24, family = "gochi"),
    axis.title.y = element_text(size=24, family = "gochi"),
    legend.position = c(0.895, 0.932),
    legend.background = element_rect(colour = "black", fill="white", size=0.5, linetype="solid"),
    legend.text =  element_text(size=14, family = "rock"),
    legend.title = element_blank(),
  )

and the plot:

[technocrat]

@juanriera

    legend.title = element_blank(),

add

    # added, lineheight should work but elusive
    axis.ticks = element_line(linewidth = 4)

[juanriera]

Thanks Richard, I can see that my English is awful. It's not the ticks what I do need but labelling all ticks. My axis was labelled (-1, --, 0, --, 1) with empty labels at minor ticks, I wanted all labels like (-1, -0.5, 0, 0.5, 1)
I think I've got it now. I couldn't imagine that getting this graph would be so complicated! Maybe there's an easier way to get the result... but I think I'm done with this by now.
Final code :

# ----------------------------------------------------------------------------
# Mi versión con ggplot
library(tidyverse)
library(showtext)
library(scales)

## Loading Google fonts (https://fonts.google.com/)
font_add_google("Gochi Hand", "gochi")
font_add_google("Schoolbell", "bell")
font_add_google("Covered By Your Grace", "grace")
font_add_google("Rock Salt", "rock")

## Automatically use showtext to render text for future devices
showtext_auto()

## Tell showtext the resolution of the device,
## only needed for bitmap graphics. Default is 96
showtext_opts(dpi = 96)

set.seed(123)
x  <- rnorm(10)
y  <-  1 + x + rnorm(10, sd = 0.2)
y[1] <-  5
mod <- lm(y ~ x)

df <- data.frame(x,y)

df |> ggplot(aes( x = x, y = y)) +
  geom_point(size = 3, colour = "steelblue") + # colores: cadetblue4, dodgerblue3, firebrick4, 
  geom_smooth(aes(colour = "black"), method = 'lm', formula = y~x, se = F) +
  geom_abline(aes(slope = 1, intercept = 1, colour = "red")) +
  scale_colour_manual(values=c("black", "red"), labels = c("OLS", "Truth")) +
  scale_x_continuous(labels = label_number(accuracy = 0.1), breaks = breaks_extended(6)) +
  labs(
    x = "x variable",
    y = "y variable",
    title = "Draw Plots Before You Fit A Regression",
  ) +
  geom_text(
    aes(x, y), 
    data = data.frame(x = -0.5, y = 4.7), 
    label = "This is the outlier",
    colour = "steelblue",
    family = "grace",
    size = 7
  ) + 
  geom_text(
    aes(x,y),
    data = data.frame(x = 1, y = 1.2), 
    label = expression(paste("True model: ", y == x + 1)),
    family = "rock",
    angle = 30,
    size = 6, 
    colour = "red" 
  ) +
  geom_text(
    aes(x,y),
    data = data.frame(x = 0, y = 2), 
    label = expression(paste("OLS: ", hat(y) == 0.79 * x + 1.49)),
    family = "rock",
    angle = 22,
    size = 7, 
    colour = "black" 
  ) +
  theme_bw() + 
  theme(
    plot.title = element_text(size = 22, family = "bell"),
    axis.text.x = element_text(size=18, family = "gochi"),
    axis.text.y = element_text(size=18, family = "gochi"),
    axis.title.x = element_text(size=24, family = "gochi"),
    axis.title.y = element_text(size=24, family = "gochi"),
    panel.grid.minor.x = element_blank(),
    legend.position = c(0.895, 0.932),
    legend.background = element_rect(colour = "black", fill="white", size=0.5, linetype="solid"),
    legend.text =  element_text(size=14, family = "rock"),
    legend.title = element_blank(),
  )

and final plot: