1. scatter plot
2. line plot
3. boxplot with significant test

1. order the points according to a particular variable

1. highlight key data points
2. insert texts and lines
3. modify the shape or color

Plotting with multiple groups or conditions

scatter plot



library(ggplot2)

# The first six rows of the data are showed below.
head(datause)
   method      theta     time
1    MFC   -1.2070657 826.5765
2    MFC    0.2774292 501.9339
3    MFC    1.0844412 607.5252
4    MFC   -2.3456977 621.7279
5    MFC    0.4291247 448.1083
6    MFC    0.5060559 725.4090

ggplot(datause, aes(x = theta, y = time))+
  geom_point(aes(color = method, shape = method))+
  scale_color_manual(values = c("#00AFBB", "#E7B800", "#FC4E07"))+

  # If you want to separate each group like I did, try the line below.
  facet_wrap(~method, ncol = 3, nrow = 1)+

  theme_set(theme_classic())+
  theme(legend.title = element_blank(),
        axis.text = element_text(size = 12, family="serif"),
        axis.title = element_text(size = 14, family="serif"),
        axis.title.x = element_blank(),
        axis.text.x = element_text(size = 15, family="serif", color = "black"),
        axis.text.y = element_text(size = 15, family="serif", color = "black"),
        plot.title = element_text(size = 14, family="serif"),
        legend.text = element_text(size = 15, family="serif"),
        strip.text = element_text(size = 15, family="serif"))
	

line plot



library(ggplot2)

# The first six rows of the data are showed below.
head(datause)
    nitems selection consistency   type  cut
1     20          RA   0.7541247  Rudner 25th
2     20       Theta   0.7671183  Rudner 25th
3     20         Cut   0.7671730  Rudner 25th
4     20          WM   0.7672108  Rudner 25th
5     20          GP   0.7671798  Rudner 25th
6     50          RA   0.8234817  Rudner 25th

plot.CC <- ggplot(data = datause,
                  aes(x = nitems, y = consistency, group = selection)) +
  geom_line(aes(colour = selection), size = 1) +
  geom_point(size = 2, aes(colour = selection, shape = selection)) +
  labs(x = "number of items", y = "Classification Consistency") +
  scale_x_continuous(breaks = c(20, 50, 70, 100)) +

  theme(text = element_text(size = 15, family = "sans"),
        axis.text.x = element_text(size = 15, face = "bold"),
        axis.text.y = element_text(size = 15, face = "bold"),
        plot.margin = unit(rep(3, 4), 'lines'),
        axis.title.x = element_text(vjust = -3),
        axis.title.y = element_text(vjust = 5),
        panel.spacing = unit(1, "lines"),
        legend.title = element_blank())

# We need to separate each plot according to the type of the index.
plot.CC + facet_grid(cut ~ type, scales = "free")

boxplot with significant test



library(ggplot2)
library(ggpubr)      # This is the main character.

# The first six rows of the data are showed below.
head(datause)
    method   PCC
1     MFC    0.922
2     MFC    0.921
3     MFC    0.923
4     MFC    0.920
5     MFC    0.922
6     MFC    0.924

# Define the group pair that you want to test.
my_comparisons <- list(c("MFC", "timed-MFC (modified)"), 
                       c("timed-MFC (modified)", "timed-SAI"), 
                       c("MFC", "timed-SAI"))

# Great! We can make a boxplot first.
ggboxplot(datause, x = "method", y = "PCC",
	color = "method", palette = c("#00AFBB", "#E7B800", "#FC4E07"),

	# Add all your data onto the plot.
	add = "jitter"
	) +
	theme(legend.title = element_blank(),
	    axis.text = element_text(size = 12, family="serif"),
	    axis.title = element_text(size = 14, family="serif"),
	    axis.title.x = element_blank(),
	    axis.text.x = element_text(size = 15, family="serif", color = "black"),
	    axis.text.y = element_text(size = 15, family="serif", color = "black"),
	    plot.title = element_text(size = 14, family="serif"),
	    legend.text = element_text(size = 15, family="serif")
	    ) +
# Next, we can use t-test for comparision.
stat_compare_means(comparisons = my_comparisons, method = "t.test")

Organizing the scatter points

order the points according to a particular variable



library(ggplot2)
library(dplyr)

# The first six rows of the data are showed below.
head(datause)
   CNY code valid       corre modelFIT
1  QAZ   31     0   0.2560892  non-fit
2  ARG   32     0   0.2759617  non-fit
3  AUS   36     1   0.3932474      fit
4  CHL  152     1   0.3015014      fit
5  COL  170     0   0.3113862  non-fit
6  CRI  188     0   0.2261491  non-fit

# This step is to create a temporary data.frame to restore the order data.
# e.g, here, I want to re-order the data according to the correlation (i.e., corre)
tempDF <-
  infludata %>% 
  arrange(corre)

head(tempDF)
   CNY code valid      corre modelFIT
1  DOM  214     0  0.1963621  non-fit
2  PAN  591     0  0.2072438  non-fit
3  CRI  188     0  0.2261491  non-fit
4  LTU  440     0  0.2500453  non-fit
5  QAZ   31     0  0.2560892  non-fit
6  ISR  376     0  0.2586625  non-fit

# Now you get the desired order, and you can start to plot.
order <- tempDF$CNY

ggplot(datause, aes(x = CNY, y = corre, color = modelFIT)) + 
  geom_point(size = 6) +
  scale_x_discrete(limits = order) +
  scale_color_manual(values = c('#132b43','#56b1f7'))+
  labs(x = "OECD countries",
       y = "Correlation between the enjoyment of reading \n and reading ability") +
  labs(color = "construct validity of\nthe enjoyment of reading")+
  theme(legend.position = "right",
        legend.title = element_text(size = 14),
        legend.text = element_text(size = 14),
        axis.text = element_text(size = 13),
        axis.title = element_text(size = 14), 
        plot.title = element_text(size = 14))

Adding guidelines or other elements

highlight key data points



library(ggplot2)

# The first six rows of the data are showed below.
head(datause)
        info         ET        value
1  0.2070762  108.69578  0.003827824
2  0.2166885   92.40194  0.005731924
3  0.2430004   76.77980  0.010955015
4  0.7163323  102.87900  0.014836777
5  0.9668220  109.88926  0.017486031
6  0.1802682   19.84861  0.005187645

# I want to order the data points according to "value" and find the top 10 points,
# so I first create a new variable "ord" to signal the desired order.
datause$ord = findInterval(datause$value, sort(datause$value))

# Then, I can get my key points.
# (You can skip this if you have known your key points)
key <- datause[order(datause$ord, decreasing = T), ][1:10, ]

# Now my data was like:
head(datause)
        info         ET        value    ord
1  0.2070762  108.69578  0.003827824     89
2  0.2166885   92.40194  0.005731924    130
3  0.2430004   76.77980  0.010955015    216
4  0.7163323  102.87900  0.014836777    257
5  0.9668220  109.88926  0.017486031    283
6  0.1802682   19.84861  0.005187645    121

# I also define a horizontal line at y-axis.
y_med <- (max(datause$info)-min(datause$info))/2

# Finally, we can start to plot.
pe <- ggplot(datause, aes(ET, info)) +
        geom_point(aes(colour = ord), size = 4) +

        # As I did here, you can add additional lines to the plot.
        geom_hline(yintercept = y_med, color = "#969696", size = 1.5, linetype = "dashed")+
        geom_vline(xintercept = 54, color = "#969696", size = 1.5, linetype = "dashed")+

        # Here, we can highlight the key points.
        geom_point(data = as.data.frame(cbind(ET = key$ET, info = key$info)), aes(ET, info),
                   color = "red", shape = 18, size = 5) +
		
        # Listed below are some codes to make the plot look prettier.
        labs(x = "expected response time", y = "information at cut score",
             title = expression(italic(v) == e^paste(mu[beta]-mu[tau]))) +
        scale_y_continuous(breaks = seq(0, 1.2, 0.2))+
        scale_color_gradient(low = "#c6dbef", high = "#08519c") +
        labs(colour="selected\norder\n")+
        theme_minimal()+
        theme(panel.grid = element_blank(),
              panel.border = element_rect(fill = NA, color="black", size = 1, linetype = "solid"),
              axis.line.x = element_line(linetype = 1, color = "black", size = 1),
              axis.line.y = element_line(linetype = 1, color = "black", size = 1),
              axis.ticks.x = element_line(color = "black", size = 1),
              axis.ticks.y = element_line(color = "black", size = 1),
              axis.title.x = element_text(margin = margin(t = 20), size = 18, family = "serif"),
              axis.title.y = element_text(margin = margin(r = 20), size = 18, family = "serif"),
              plot.title = element_text(margin = margin(b = 20), hjust = 0.5, size = 18, family = "serif"),
              legend.position = "right",
              legend.title = element_text(size = 14, family = "serif"),
              axis.text = element_text(size = 15, color = "black", family = "serif"),
              legend.text = element_text(size = 13, family = "serif"))
pe + coord_fixed(ratio = 160, xlim = c(0, 200))
								

insert texts and lines



library(ggplot2)
library(ggridges)
library(hrbrthemes)

# The first six rows of the data are showed below.
head(datause)
     th       lnL
1 -3.00 -74.94047
2 -2.99 -74.64811
3 -2.98 -74.35579
4 -2.97 -74.06350
5 -2.96 -73.77125
6 -2.95 -73.47904

# This is our base plot, just a parabola.
p <- ggplot(datause, aes(x = th, y = lnL)) +
  geom_line(color = "black", size = 1.5, alpha = 0.9, linetype = 1) +
  scale_x_continuous(breaks = seq(-3, 3, 0.5), labels = seq(-3, 3, 0.5)) +
  scale_y_continuous(breaks = seq(-80, 10, 10), labels = seq(-80, 10, 10)) +
  theme_ipsum() +
  theme(axis.text.x = element_text(size = 15, family="serif", color = "black"),
        axis.text.y = element_text(size = 15, family="serif", color = "black")) +
  xlab(" ") + 
  ylab(" ") +
  ggtitle(" ") +

# Next, you can include some guidelines on the x-axis and y-axis using different colors.
# The setting of value is totally up to you.
  geom_segment(x = -0.4, xend = -0.4, y = -100, yend = datause$lnL[261],
               color = "blue", linetype = 2, size = 1)+
  geom_segment(x = -0.6, xend = -0.6, y = -100, yend = datause$lnL[241],
               color = "blue", linetype = 2, size = 1)+
  geom_segment(x = datauseth[which.max(datauseth[which.max(datauselnL)], xend = datauseth[which.max(datauseth[which.max(datauselnL)],
               y = -100, yend = max(datause$lnL),
               color = "blue", linetype = 2, size = 1)+
  geom_segment(x = -0.4, xend = -4, y = datauselnL[261],yend=datauselnL[261], yend = datauselnL[261],
               color = "red", linetype = 1, size = 1)+
  geom_segment(x = -0.6, xend = -4, y = datauselnL[241],yend=datauselnL[241], yend = datauselnL[241],
               color = "red", linetype = 1, size = 1)+
  geom_segment(x = datauseth[which.max(datauseth[which.max(datauselnL)], xend = -4,
               y = max(datauselnL),yend=max(datauselnL), yend = max(datauselnL),
               color = "red", linetype = 1, size = 1)

# You are encouraged to define the texts you want to add and their locations before the next step.
annotation1 <- data.frame(x = c(-3, -3, -3),
			  y = c(datauselnL[261],datauselnL[261], datauselnL[241], max(datause$lnL)),
			  label = c("-9.68", "-12.41", "-6.60"))
annotation2 <- data.frame(x = c(-0.7, -0.3, datauseth[which.max(datauseth[which.max(datauselnL)]),
			  y = c(-33, -33, -33),
			  label = c("(SPRT)", "(SPRT)", "(GLR)"))

# The merits of this is obvious.
# You can easily include those annotations through geom_label.
p + geom_label(data = annotation1, aes(x = x, y = y, label = label),
               size = 5, family = "serif",
               label.size = NA) +
    geom_label(data = annotation2, aes(x = x, y = y, label = label),
               size = 5, family = "serif",
               label.size = NA) +

# But life isn't always smooth sailing,
# even so, if you also need to label some formulae, you can add them in this way.
 annotate("label", x = -0.7, y = -30, label = expression(italic(theta[l])),
          size = 6, family = "serif", label.size = NA)+
 annotate("label", x = -0.3, y = -30, label = expression(italic(theta[u])),
          size = 6, family = "serif", label.size = NA)+
 annotate("label", x = datauseth[which.max(datauseth[which.max(datauselnL)], y = -30,
          label = expression(italic(theta[1])),
          size = 6, family = "serif", label.size = NA)+

# The codes below are used to make three points stand out.
# If you do not need to do so, feel free to skip this part.
 annotate("point", x = -0.6, y = datause$lnL[241], size = 5)+
 annotate("point", x = -0.4, y = datause$lnL[261], size = 5)+
 annotate("point", x = datauseth[which.max(datauseth[which.max(datauselnL)], y = max(datause$lnL), size = 5)+
 annotate("point", x = -0.6, y = datause$lnL[241], size = 3, color = "#ffffff")+
 annotate("point", x = -0.4, y = datause$lnL[261], size = 3, color = "#ffffff")+
 annotate("point", x = datauseth[which.max(datauseth[which.max(datauselnL)], y = max(datause$lnL), size = 3, color = "#ffffff")

								

modify the shape or color



library(ggplot2)

# The first six rows of the data are showed below.
head(datause)
   theta    PCC method
1  -3.50 0.0017    MFC
2  -3.25 0.0021    MFC
3  -3.00 0.0022    MFC
4  -2.75 0.0029    MFC
5  -2.50 0.0022    MFC
6  -2.25 0.0038    MFC

# If your label name includes formulae, defining it as a variable will make things easier.
labelname <- c(expression(PCC[MFC] - PCC[Random]), expression(PCC[SAI] - PCC[Random]))

# Now we can start to plot.
w05 <- ggplot(datause)+
  geom_point(aes(x = theta, y = PCC, shape = method, col = method, alpha = method),
             size = 5, stroke = 1.5)+
  geom_hline(yintercept = 0, size = 1.3, linetype = "dashed")+

  # Here is the point, you can define different shapes and colors for different groups.
  scale_shape_manual(values = c(5, 16),
                     labels = labelname)+
  scale_color_manual(values = c("blue", "red"),
                     labels = labelname)+
  scale_alpha_manual(values = c(0.8, 0.5),
                     labels = labelname)+

  # Listed below are codes to limit the range of x-axis and y-axis.
  scale_y_continuous(limits = c(-0.2, 0.2), breaks = seq(-0.2, 0.2, 0.1))+
  scale_x_continuous(breaks = seq(-3.5, 3.5, 1))+

  # Listed below are some codes to make the plot look prettier.
  labs(x = "theta point", y = "PCC differences",
       title = bquote(paste(italic(w), " = 0.50, ", italic(J) == .(Ji))))+
  theme_minimal()+
  theme(panel.grid = element_blank(),
        panel.border = element_rect(fill = NA, color="black", size = 1, linetype = "solid"),
        axis.line.x = element_line(linetype = 1, color = "black", size = 1),
        axis.line.y = element_line(linetype = 1, color = "black", size = 1),
        axis.ticks.x = element_line(color = "black", size = 1),
        axis.ticks.y = element_line(color = "black", size = 1),
        axis.title.x = element_text(margin = margin(t = 20), size = 30, family = "serif"),
        axis.title.y = element_text(margin = margin(r = 20), size = 30, family = "serif"),
        plot.title = element_text(margin = margin(b = 20), hjust = 0.5, size = 30, family = "serif"),
        axis.text = element_text(size = 25, color = "black", family = "serif"),
        
        legend.position = c(.78, .12),
        legend.text = element_text(size = 20, family = "serif", hjust = 0),
        legend.title = element_blank())
w05 + coord_fixed(ratio = 12)
								



library(ggplot2)

# The first six rows of the data are showed below.
head(datause)
       type       theta
1 estimated  0.02181399
2 estimated  0.50863802
3 estimated  0.55536742
4 estimated -0.09117160
5 estimated -0.60815771
6 estimated -0.66477364

ggplot(data = datause, aes(x = theta, group = type, fill = type)) +

  # You can modify the color and transparency here.
  geom_density(adjust = 1.5, alpha = 0.4, size = 0.75) +

  theme(text = element_text(size = 15, family = "sans"),
        axis.text.x = element_text(size = 15),
        axis.text.y = element_text(size = 15),
        legend.title = element_blank())