I am in the middle of reading "Code Complete" and find it quite nice, but don't know how I can directly translate this to my code in a helpful way.
If you're writing one algorithm to solve one problem, it's enough to make it correct and fast, and you probably don't need to care about DRY, SOLID, YAGNI or anything else. But if it's a "real world" project, these concepts can help you change course quickly when requirements change.
FWIW I don't remember anything I read in Code Complete. Maybe it's time to skim it again.
Repeat.