Want to share your content on R-bloggers? click here if you have a blog, or here if you don't.

Generating violin boxplots from image. Yes, why not.

We create a raster image from a picture and calculating the ratio of the pixels on the scale of grayscale. The more the darker colour is represented in the pixels, the bigger the value. And this value is converted into the vector of values. And each vector is represneted as a violin boxplot.

```##########################################
#
# Converting JPG and plots raster using horizontal violins
#
# Series:
# Little Useless-useful R functions #33
# Created: January 29, 2022
# Author: Tomaz Kastrun
# Blog: tomaztsql.wordpress.com
# V.1.0
###########################################

# Libraries
pkg <- c("dplyr", "tidyr", "ggplot2", "magick", "stringr",
"forcats", "viridis", "grid", "purrr","hrbrthemes")
lapply(pkg, require, character.only = TRUE)

img <- img %>%
image_quantize(max=2, colorspace = 'gray', dither=TRUE) %>%
image_scale(geometry = geometry_size_pixels(width=50, height=15, preserve_aspect=FALSE))

# Image manipulation
mat <- t(1L - 1L * (img[[1]][1,,] > 180))
mat_df <-data.frame(mat)

# Transpose  data
dff <- data.frame(x = NULL, y = NULL)
for (i in 1:nrow(mat_df)) {
for (j in 1:ncol(mat_df)){
if (mat_df[i,j] == 1){
d <- data.frame(x=i, y=j)
dff <<- rbind(dff, d)
}}}

# Creating factors

# Reversing order
df <- dff %>%
mutate(x = fct_rev(fct_reorder(x,y))) %>%
purrr::map_df(rev)

df2 <- dff %>% mutate(x = fct_rev(fct_reorder(x,y)))
df3 <- data.frame(x = as.character(df2\$x), y = df\$y)

vp <- df3 %>%
ggplot( aes(x=x, y=y, fill=x, color=y)) +
geom_violin(width=1.5, size=0.2) +
scale_fill_viridis(discrete=TRUE) +
scale_color_viridis(discrete=TRUE) +
theme_void()

print(vp, vp=viewport(angle=-90))  # flip and change graph orientation
```

With the library imager, I will also prepare an additional post no better calculation of greyscale and converting them to boxplots.

As always, code is available in at the Github in same Useless_R_function repository. Check Github for future updates.

Happy R-coding and stay healthy!“