The paper is the RatInattPlus.pdf file in this directory. The other pdf file is a set of slides from a talk at Princeton based on that paper.

The R files are those used to solve the paper's two-period saving problem. You can use them to experiment with other utility functions U, other distributions g of wealth, or different discretizations from those used in the paper.

screpD:

Calculates the objective function (expected utility minus lambda times mutual information between W and C) for a given f(c,w) joint pdf. Returns a list with much useful information.

screpDCS

Calls screpD, then strips everything from the list except function value, reverses the sign on that, and returns it. This is the kind of function csminwel.R expects.

makeGraphs

computes p(w|c) and p(c), makes grey-shaded contour graphs of p(w|c), p(c|w), and joint pdf f(c,w) and a line graph of p(c). Conditional distribution charts may need to be cleaned up to get rid of garbage corresponding to zero-probability conditioning values. The p(c) charts in the paper were done with lty="h", lwd=10, which are not implemented in this program.