Selecting right base graph

principles
Rcode
example
Multiple examples and code from the graphics principles cheatsheet section: selecting the right base graph.
Author

Graphics principles

Published

December 19, 2023

Consider if a standard graph can be used by identifying suitable designs based on the: (i) purpose (i.e. message to be conveyed or question to answer) and (ii) data (i.e. variables to display). In this post we provide multiple examples and code from the graphics principles cheatsheet section: selecting the right base graph.

Deviation (i.e. change from baseline)

Example graphs for displaying deviations from a common reference point or baseline include line plots displaying change from baseline or a diverging barchart such as a waterfall plot.

# generate data and plot with a minimal theme
tibble(
  trt = rep(c("Dose 1", "Dose 2", "Dose 3"), each = 4),
  visit = rep(c(1, 2, 3, 4), 3),
  response = c(0, 0.7, 0.8, 0.6,
               0, 0.5, 0.6, 0.2,
               0, -0.2, -0.6, -0.9)
) %>%
  ggplot(aes(x = visit, y = response, group = trt)) +
  geom_hline(
    yintercept = 0,
    colour = "wheat4",
    linetype = 1,
    size = 0.6
  ) +
  geom_point(size = 3) +
  geom_line(size = 1) +
  scale_y_continuous(limits = c(-1, 1), breaks = c(-1,-0.5, 0, 0.5, 1)) +
  theme(
    legend.position = "none",
    axis.title.y = element_blank(),
    axis.text.y = element_blank(),
    axis.title.x = element_blank(),
    axis.text.x = element_blank(),
    panel.grid.minor = element_blank(),
    panel.grid.major = element_blank()
  )
Warning: Using `size` aesthetic for lines was deprecated in ggplot2 3.4.0.
ℹ Please use `linewidth` instead.

Line plot displaying change over time from a fixed baseline. This is often called a spaghetti or patient profile plot.
## generate data
tibble(x = 1:10,
       ratio = sort(runif(10, 0, 2))) %>%
  mutate(col_flag = if_else(ratio > 1, 1, 0)) %>%
  ggplot(aes(color = factor(col_flag))) +
  geom_segment(aes(
    x = x,
    xend = x,
    y = 1,
    yend = ratio
  ), size = 8) +
  geom_hline(
    yintercept = 1,
    colour = "wheat4",
    linetype = 1,
    size = 1
  ) +
  ggColor(color.hues = c(2, 1)) +
  theme(
    panel.grid.minor = element_blank(),
    panel.grid.major = element_blank(),
    axis.ticks = element_blank(),
    axis.text = element_blank(),
    legend.position = "none",
    axis.title.x = element_blank(),
    axis.title.y = element_blank()
  )

Waterfall or diverging bar chart.

Correlation

For displaying the relationship between them two or more variables, x-y displays such as scatter plots and heatmaps are useful starting points.

tibble(cause = c(runif(40, 1, 10)),
       effect = cause * (rnorm(40, 2, 0.5)) + rnorm(40, 0, 1)
       ) %>%
  ggplot(aes(x = cause, y = effect)) +
  geom_point(size = 4.5,
             shape = 16,
             position = "jitter") +
  geom_smooth(method = "lm") +
  scale_y_continuous(expand = c(0, 0.5)) +
  scale_x_continuous(expand = c(0, 0.5)) +
  theme(
    panel.grid = element_blank(),
    legend.position = "none",
    axis.title.x = element_blank(),
    axis.text.x = element_blank(),
    axis.title.y = element_blank(),
    axis.text.y = element_blank()
  )
`geom_smooth()` using formula = 'y ~ x'

Scatter plot with a regression line
mydata <- mtcars[, c(1, 3, 4, 5, 6, 7)]
cormat <- round(cor(mydata), 2)
melted_cormat <- melt(cormat)

# Get lower triangle of the correlation matrix
get_lower_tri <- function(cormat) {
  cormat[upper.tri(cormat)] <- NA
  return(cormat)
}

# Get upper triangle of the correlation matrix
get_upper_tri <- function(cormat) {
  cormat[lower.tri(cormat)] <- NA
  return(cormat)
}

upper_tri <- get_upper_tri(cormat)
melted_cormat <- melt(upper_tri, na.rm = TRUE)

reorder_cormat <- function(cormat) {
  # Use correlation between variables as distance
  dd <- as.dist((1 - cormat) / 2)
  hc <- hclust(dd)
  cormat <- cormat[hc$order, hc$order]
}

# Reorder the correlation matrix
cormat <- reorder_cormat(cormat)
upper_tri <- get_upper_tri(cormat)

# Melt the correlation matrix
melted_cormat <- melt(upper_tri, na.rm = TRUE)

# Create a ggheatmap
ggplot(melted_cormat, aes(Var2, Var1, fill = value)) +
  geom_tile(color = "white") +
  scale_fill_gradient2(
    low = Novartis.Color.Palette[1, 1],
    high = Novartis.Color.Palette[1, 2],
    mid = "white",
    midpoint = 0,
    limit = c(-1, 1),
    space = "Lab"
  ) +
  coord_fixed() +
  guides(fill = guide_colorbar(barwidth = 7, barheight = 1)) +
  theme(
    axis.text = element_blank(),
    axis.title.x = element_blank(),
    axis.title.y = element_blank(),
    legend.title = element_blank(),
    panel.grid = element_blank(),
    panel.border = element_blank(),
    panel.background = element_blank(),
    axis.ticks = element_blank(),
    legend.justification = c(1, 0),
    legend.position = c(0.6, 0.7),
    legend.direction = "horizontal"
  )

Correlation heatmap.

Ranking

Bar charts and dot plots are effective for disaplying quantities ordered highest to lowest (or vice versa).

my_data <- data.frame(
  grp = c("A", "B", "C", "D"),
  perc = c(1, 0.8, 0.6, 0.2)
)

ggplot(my_data, aes(x = perc, y = reorder(grp, perc))) +
  geom_point(size = 6) +
  scale_x_continuous(breaks = seq(0, 1, 0.2),limits = c(0, 1)) +
  theme_minimal() +
  theme(panel.grid.major.y = element_line(colour = "grey60", size = 0.8),
        panel.grid.major.x = element_blank(),
        panel.grid.minor = element_blank(),
        axis.title.x=element_blank(),
        axis.text.x=element_blank(),
        axis.title.y=element_blank(),
        axis.text.y=element_blank()
  )
Warning: The `size` argument of `element_line()` is deprecated as of ggplot2 3.4.0.
ℹ Please use the `linewidth` argument instead.

df <- data.frame(trt=c("A", "B", "C","D","E"),
                 cause=c(1,2,4,6,10),
                 highlight = c(2,1,2,2,2))


ggplot(df, aes(x=trt, y=cause, group=trt)) + theme_minimal(base_size=18) +
  geom_bar(width=0.7,fill=Novartis.Color.Palette[1,1], stat = "identity") +  
  scale_y_continuous(breaks=c(0, 5, 10)) + 
  geom_hline(yintercept = 0, colour = "wheat4", linetype=1, size=1)+ 
  coord_flip() +
  theme(legend.position="none",
        panel.grid.minor.x = element_blank(),
        panel.grid.minor.y = element_blank(),
        panel.grid.major.y = element_blank(),
        axis.ticks = element_blank(),
        axis.title=element_blank(),
        axis.text=element_blank()
  )

Distribution

Histograms, density plots, boxplots and violin plots are common displays for visualising the distribution of a variable.

### dummy data generation
set.seed(131)

mu1 <- 135.5 # population mean; 
sd1 <- 5.5 # sd of population mean across trials
NVS301  <- rnorm(200, mu1, sd1) 

mu2 <- 140.1 # population mean; 
sd2 <- 5.9 # sd of population mean across trials
NVS201  <- rnorm(200, mu2, sd2) 

mu3 <- 145.5 # population mean; 
sd3 <- 7 # sd of population mean across trials
SOC101  <- rnorm(200, mu3, sd3)

bp <- c(NVS301,NVS201,SOC101)

trt <- 1:600
trt[1:200] = "NVS301"
trt[201:400] = "NVS201"
trt[401:600] = "SOC101"

trtn <- 1:600
trtn[1:200] = 1
trtn[201:400] = 2
trtn[401:600] = 3


# Put in to simple data frame 
dat1 <- data.frame( 
  treatment = factor(trt, levels = c("NVS301", "NVS201", "SOC101")),
  bpm = bp
) 

ggplot(dat1, aes(x=as.factor(treatment), y=bpm, fill=treatment)) + 
  geom_violin(trim=TRUE)+ggFill(color.values=c("Tint1"))+
  geom_boxplot(width=0.2, fill="white")+
  #scale_fill_brewer(palette="RdBu") +
  theme_minimal() +
  theme(legend.position='None',
        panel.grid.minor = element_blank(),
        panel.grid.major = element_blank(),
        axis.ticks.x = element_blank(),
        axis.title=element_blank(),
        axis.text=element_blank(),
        panel.background = element_blank()
  )  

set.seed(1234)
dat <- data.frame(cond = factor(rep(c("A","B"), each=200)), 
                  rating = c(rnorm(200),rnorm(200, mean=.8)))


# Histogram overlaid with kernel density curve
ggplot(dat, aes(x=rating)) +
  geom_histogram(aes(y=..density..), binwidth=.4,colour="black", fill=Novartis.Color.Palette[1,1]) +
  geom_density(alpha=.1, color = "gray", fill=Novartis.Color.Palette[1,1])+
  geom_hline(yintercept = 0, colour = "wheat4", linetype=1, size=1)+ 
  theme_minimal() +
  theme(legend.position='None',
        panel.grid.minor = element_blank(),
        panel.grid.major = element_blank(),
        axis.ticks = element_blank(),
        axis.title=element_blank(),
        axis.text=element_blank(),
        panel.background = element_blank(),
        panel.border = element_blank()
  )
Warning: The dot-dot notation (`..density..`) was deprecated in ggplot2 3.4.0.
ℹ Please use `after_stat(density)` instead.

Change over time (i.e. evolution)

Line plots can be used for displaying how quantities evolve over time.

fit <- survfit(Surv(time, status) ~ sex, data = lung)
gg<-autoplot(fit, conf.int = FALSE, censor = FALSE, surv.size = 2)
gg + theme_minimal() + ggColor() + 
  theme(legend.position='None',
        axis.ticks = element_blank(),
        axis.title=element_blank(),
        axis.text=element_blank(),
        panel.grid.minor = element_blank(),
        panel.grid.major = element_blank(),
        panel.background = element_blank()
  ) 

#make data
df <- data.frame(trt=rep(c("Dose 1", "Dose 2", "Dose 3"), each=4),
                 visit=rep(c(1, 2, 3, 10),3),
                 response=c(0.8, 2.9, 4.2,2.5, 0.4, 2.4,3.8,1.0,0.2, 2.1,3.0,0.2))
#head(df)

# define a minimal theme
th_con <- theme_minimal(base_size = 12 ) +  
  theme(legend.position="none",
        axis.title.y=element_blank(),
        axis.text.y=element_blank(),
        axis.title.x=element_blank(),
        axis.text.x=element_blank(),
        panel.grid.minor=element_blank(), 
        panel.grid.major=element_blank()
  )    



ggplot(df, aes(x=visit, y=response, group=trt, color = trt)) + 
  geom_line(size=1.5) + ggColor() + 
  scale_y_continuous(limits = c(0, 5), breaks = c(0,2.5, 5)) +
  th_con

Part-to-whole

For displaying sub‐divisions of a whole (e.g. the percentage of patients in a subgroup), pie charts, bar charts and stacked barcharts can be used.

# Generate data
df <- data.frame(trt=c("A","A","B", "B","C", "C"),
                 subgroup = c("1","2","1","2","1","2"),
                 percent=c(0.5,0.5,0.4,0.6,0.2,0.8)
)  

ggplot(df, aes(x = factor(trt), y = percent, fill = subgroup)) +
  geom_bar(width=0.6, stat="identity") + 
  scale_y_continuous(breaks=c(0, 0.5, 1)) + ggFill() + 
  #   scale_fill_brewer(palette="Blues") +
  geom_hline(yintercept = 0, colour = "wheat4", linetype=1, size=1)+ 
  theme_minimal(base_size=9) +
  theme(panel.grid.minor=element_blank(), 
        panel.grid.major=element_blank(),
        panel.border=element_blank(),
        axis.ticks = element_blank(), 
        axis.text = element_blank(),
        legend.position="none",
        axis.title.x=element_blank(),
        axis.title.y=element_blank())

data(business)
treemap(business[business$NACE1=="C - Manufacturing",], 
        index = c("NACE2","NACE3"), vSize = c("employees"), vColor = c("employees"),
        type = "index", palette = Novartis.Color.Palette[,1], fontsize.labels = 0, fontsize.title = 0)

Magnitude

Dotplots, forest plots and barcharts are useful for displaying comparisons of size or magnitude (i.e. treatment differences)

df <- data.frame(trt=c("A", "B", "C","D","E"),
                 cause=c(4,6,10,12,15),
                 highlight = c(2,2,2,1,2))

ggplot(df, aes(x=c(4,1,5,2,3), y=cause, group=trt, fill=trt)) +
  geom_bar(width=0.5, stat = "identity") +  
  geom_hline(yintercept = 0, colour = "wheat4", linetype=1, size=0.8)+ 
  ggFill(color.values = "Hue", color.hues = 1) +
  scale_y_continuous(breaks=c(0, 5, 10)) + 
  theme(panel.grid.minor=element_blank(), 
        panel.grid.major=element_blank(),
        axis.ticks = element_blank(), 
        axis.text.y = element_blank(),
        axis.text.x = element_blank(),
        legend.position="none",
        axis.title.x=element_blank(),
        axis.title.y=element_blank())

theme_set(theme_minimal(base_size=26))
th <- theme(axis.title.x=element_blank(),
            axis.title.y=element_blank(),
            axis.text.x=element_blank(),
            axis.text.y=element_blank(),
            panel.grid.minor=element_blank(), 
            panel.grid.major=element_blank()
)


align <- data.frame(x=factor(c(1,2,3)),y=c(2,1,3),low=c(1,0,2),hi=c(3,2,4))
ggplot(align, aes(x=x, y=y, ymin=low, ymax=hi)) + 
  geom_point(size=7) + 
  geom_linerange(size=1) + 
  th + 
  coord_flip()