One of the more interesting applications of algorithms to images is so-called genetic art. In this technique, we build some form of image DNA, generate images from that DNA, rank the fitness of the images, and recombine the DNA, of images the DNA of more fit images more frequently in building the next generation. Electric Sheep are one of the more famous examples of genetic art. Of course, the first part of genetic art is figuring out how to turn the DNA into an image. A spectacular variety of techniques exist for such transformations. Dr. Davis has one famous early example (done by hand, rather than by computer), in which Francois Morellet used a phonebook as the DNA, and rendered it in a large grid using red for odd digits and blue for even digits. We'll consider two versions of this repurposing of data: One in which we use this same grid technique to turn a file into a series of pixels (or blocks of colors or ...) and another one in which we turn a file into a series of instructions to a turtle. We've already seen a bit of how we can interpret any file as a series of colors: We could read triplets of characters and treat each triplet as a new color. Here's one implementation of that technique, using image-calculate-pixels!, a slower variant of image-compute-pixels! that traverses the image systematically, row by row, from top-left to bottom right. One problem with that technique is that it provides a fairly wide variety of colors, giving no form at all to the image. It also doesn't scale appropriately. A simpler technique is to choose a smaller palette of related colors and use the characters as indices into that palette. That is, we pick sections of the image to draw (most simply, a grid of rectangles or ovals), and, for each section, read a character, convert the character to an integer, reduce that integer to the range of valid indices, and then get it from the palette. We might use such a procedure to make a wide variety of images by selecting palettes, files, horizontal boxes, and vertical boxes from n. While these grid-like visualizations of files are fun, they are certainly not the only thing we can do to interpret a file as the DNA of a drawing. As mentioned above, we can treat a file as a series of commands to a turtle. Unfortunately, this procedure is a bit less refined than most. Still, it may be worth exploring.