Driven IFS and Data Analysis

IFS Driven by Letters

Here is the IFS driven by the text of Benoit Mandelbrot's essay "Mathematics and Society in the Twentieth Century," read as a DNA string. How was this done?
* We took the text of the essay, and removed all punctuation, spaces, paragraph indents, and converted all letters to lower case, obtaining a string of 12,325 characters.
* Then we read through the string sequentially, applying T1 for each occurrence of c, T2 for each occurrence of a, T3 for each occurrence of t, and T4 for each occurrence of g.
The resulting driven IFS is shown on the left. We certainly see a clustering of points along the diagonal, but does that reflect any more than the greater abundance of a and t? (There are 461 c, 1080 a, 1226 t, and only 192 g.)
 
CATG picture Whole text (WT) picture
Another approach is to divide the letters of the alphabet into four bins. On the right above is the driven IFS from the same text, plotted this way:
apply T1 for every a, b, c, d, e, f, g, (4119 occurrences)
apply T2 for every h, i, j, k, l, m, (2505 occurrences)
apply T3 for every n, o, p, q, r, s, (3526 occurrences)
apply T4 for every t, u, v, w, x, y, z (2175 occurrences)
Of course, there is a natural sequential ordering of the real numbers, whereas the ordering of the alphabet is arbitrary. (Glance at your computer keyboard if you think the ordering of the letters is anything other than arbitrary.) So we must be careful with interpretations.
However, comparing these pictures can lead to the discovery of a delicate point about driven IFS.

Return to IFS Driven by Texts.