Code for making fourier transform mistakes and producing plots for weird wavy patterns.

accidental_aRt.R

library(ggplot2)

accidental_art <- function(x, r = 1.9) {
  x <- cor(x)
  n <- dim(x)[1]
  mask <- matrix(0, ncol(x), nrow(x))
  rad <- n/2
  xcen <- 0
  ycen <- 0
  
  r <- n / r
  mask <- ((row(mask) - xcen)^2 + (col(mask) - ycen)^2) <= rad^2
  mask <- mask + (((row(mask) - r)^2 + (col(mask) - r)^2) <= rad^2)
  mask <- mask + (((row(mask) - 0)^2 + (col(mask) - r)^2) <= rad^2)
  mask <- mask + (((row(mask) - r)^2 + (col(mask) - 0)^2) <= rad^2)
  
  y <- fft(x)
  yim <- asinh(Re(y))
  yim <- sinh(yim * mask)
  yim <- complex(real = yim, imaginary = Im(y))
  yim <- matrix(yim, ncol = ncol(x))
  
  z <- fft(yim, inverse = T)/length(yim)
  
  df <- reshape2::melt(Re(z))
  
  ggplot(df, aes(x = Var1, y = Var2, fill = value^2)) +
    geom_raster() +
    scale_fill_gradientn(colours = c('white', '#f5a623', '#d0021b', 'black'),
                         limits = c(0, NA), oob = scales::oob_squish, trans = "sqrt") +
    scale_y_continuous(trans = "reverse")
}

accidental_art(t(iris[, 1:4]))