At work I use ggplot2 almost for everything. I really like the mid term between high level (highcharts) and low-level (like d3 for example). The deafult theme for ggplot it’s good, and really good if you compare with the old looking R base graphics, and there is more: the ggthemes package which have some themes for ggplot objects. However, I miss the elegant and modern touch, for example in highcharts.

So I decide to play around the theme function to replicate the look and feel of highcharts. The main tasks were:

1. Change the font to a more modern one.
2. Remove grid lines (minor ones).
3. Use a more plain color palette.
4. Reduce the width of bars.
5. Put a white background.

The chosen font was Open Sans (an ultra used font nowadays) which you can use in R with extrafont package.

In the next images we can see some samples of this theme.



As you can see, the plot look more clean without the gridlines and the background. This cause less confusion (and maybe less detail) because generate more space.

Finally,

In my humble option, it look great. What do you think? Here the R code to do the magic ;). The first part defines the theme and colors, the second one we make the plots.

data(diamonds)
data <- subset(diamonds, color %in% c("E", "F", "G") & cut %in% c("Ideal", "Premium", "Good"))
data$indicator <- ifelse(data$color %in% c("G" ), 1, 0)
colors_hc <- c("#7CB5EC", "#313131", "#F7A35C",
               "#90EE7E", "#7798BF", "#AAEEEE",
               "#FF0066", "#EEAAEE", "#55BF3B",
               "#DF5353", "#7798BF", "#AAEEEE")

font_import(paths="~/Downloads/open-sans/")
loadfonts()
fonts()

theme_hc <- function(){
  theme(
    text                = element_text(family="Open Sans", size = 10),
    title               = element_text(hjust=0), 
    axis.title.x        = element_text(hjust=.5),
    axis.title.y        = element_text(hjust=.5),
    panel.grid.major.y  = element_line(color='gray', size = .3),
    panel.grid.minor.y  = element_blank(),
    panel.grid.major.x  = element_blank(),
    panel.grid.minor.x  = element_blank(),
    panel.border        = element_blank(),
    panel.background    = element_blank(),
    legend.position     = "bottom",
    legend.title        = element_blank()
  )
}

p1 <- ggplot(data) +
  geom_bar(aes(cut), width =.4, fill = colors_hc[1]) +
  ggtitle("An interesting title for a bar plot") +
  xlab("Cut") + ylab("Amount") +
  scale_y_continuous(labels = comma) +
  theme_hc()
p1

p2 <- ggplot(data) +
  geom_bar(aes(color, fill=cut), position="dodge", width=.4) +
  ggtitle("Another interesting title") +
  xlab("Cut") + ylab("Amount") +
  scale_y_continuous(labels = comma) +
  scale_fill_manual(values=colors_hc) +
  theme_hc()
p2


p3 <- ggplot(data) + geom_density(aes(x, fill=cut, color=cut), alpha=I(0.5)) +
  ggtitle("Density plot") +  xlab("x") + ylab("Density") +
  scale_y_continuous(labels = percent) +
  scale_fill_manual(values=colors_hc) +
  xlim(4, 8) +
  theme_hc()
p3

Bonustrack: More is less! This is a good lesson from Darkhorse Analytics.