How about some colours in line graph?
Or even more wacky? Nevertheless, let’s create a function that generates some sample “mocked” data and draws a line chart:
#The function
Colourful_graph <- function(n, x, y){
df <- data.frame(x=x, y=y, col=n)
for (i in 1:n){
#get last x,y
lastx <- tail(df$x, 1)
lasty <- tail(df$y, 1)
col <- sample(1:i, 1, replace = T)
if (i %% 10 == 0) {
xx <- runif(1, 0.0, 1.0) + lastx
yy <- runif(1, 0.0, 1.0) - lasty
} else {
xx <- runif(1, 0.0, 1.0) + lastx
yy <- runif(1, 0.0, 1.0) - lasty
}
# change: col=i for rainbow colours
df <- rbind(df, c(x=xx, y=yy, col=col))
}
fake <- df
brewColours <- as.integer(length(fake$col))
ColourfulColours <- colorRampPalette(brewer.pal(8, "Set2"))(brewColours)
# show faked graph
ggplot(fake, aes(x, y, color = factor(col))) +
geom_line(color= ColourfulColours) +
theme(legend.position = "none") +
theme(panel.grid.major = element_blank(), panel.grid.minor = element_blank(),
panel.background = element_blank(), axis.line = element_blank(), text=element_blank(), line = element_blank())
}
And run all together with three parameters: n – number of data points and x,y for starting point.
#create colourful graph
Colourful_graph(500,0.4,0.3)
In addition, adding some colorbrewer functionalities can also make line chart random with colours. This following section can be added to function to alternate between the colour palettes:
#brewer sample 8 color palette
nn <- as.data.frame(brewer.pal.info)
nn$names <- row.names.data.frame(nn)
nn8 <- subset(nn, maxcolors==8, select=c(names))
brewColours <- as.integer(length(fake$col))
#ColourfulColours <- colorRampPalette(brewer.pal(8, "Paired"))(brewColours)
ColourfulColours <- colorRampPalette(brewer.pal(8, sample(nn8$names,1,T)))(brewColours)
Couple of more faked dataset line charts:
As always, code is available in at the Github in same Useless_R_function repository.
Happy R-coding!
TomazTsql 
[…] Tomaz Kastrun has fun with colors: […]
LikeLike
[…] by data_admin [This article was first published on R – TomazTsql, and kindly contributed to R-bloggers]. (You can report issue about the content on this page […]
LikeLike