Performance and Profiling in Racket | Schema Programming Part 30

Racket is generally fast (JIT-compiled), but you can make it faster. The first step is always measurement.

The Profiler

Don't guess where your code is slow. Use the profile library.

#lang racket
(require profile)

(define (fib n)
  (if (< n 2) n (+ (fib (- n 1)) (fib (- n 2)))))

(profile (fib 30))

This prints a report showing which functions took the most time and how many times they were called.

Typed Racket (Again)

We mentioned Typed Racket in Part 13. Aside from safety, it is also the primary performance tool. The compiler can use type information to unbox numbers and inline operations.

Fixnum/Flonum Optimizations

If you are doing heavy math in standard Racket and need raw speed, use specific operations for "Fixnums" (small integers) and "Flonums" (floating point numbers).

(require racket/fixnum)

(fx+ 10 20) ; Faster than + because it skips overflow checks

Warning: These operations are unsafe if you overflow!

Vector vs List

Lists are great for recursion, but they have O(N) access time. For performance-critical code, use Vector for O(1) access.

(define v (vector 1 2 3))
(vector-ref v 2) ; Instant

Summary

1. Profile first.
2. Algorithm change second (loops vs recursion).
3. Data structure change third (Vector vs List).
4. Typed Racket / Fixnum ops last.