I currently maintain a score of around 400 problems solved at Project Euler. It may be great or not so great, depending on whom you ask, but I believe it's safe to say I'm not a newb at CS and coding - and I never brute-force and always follow the under-one-minute-runtime rule!
And yet every time I solve one there and thus open a thread with solutions, I often bump into solutions that look like small miracles to me, which makes me wonder whether I'm missing some great gems of CS and computational math. (For those who are unfamiliar with Project Euler, there's a closed discussion to every problem, which opens only when you post the right answer.)
These threads are great sources themselves, but I'm looking for something bigger that gives consistent perspective on a topic or a number of topics. A book, basically. For example, recently I've discovered "Prime numbers, computational perspective", and it already helped me with PE, though it's been a week or so since I discovered it.
What would you recommend of this sort? Exploration of the exotic areas of CS or certain unorthodox perspectives on the not-so-exotic areas of CS which tend to be missed even by more-or-less experienced at CS and coding.
EDIT: I was asked about my background and my motivation in the comments, so here they are.
It's a hobby and I want to be better at solving puzzles, yes. Shouldn't mean I don't take it seriously and it's just about finding an algorithm on the web and copying it without any understanding of how it works for me...
I definitely don't want to be a CS researcher, but I don't mind going through something you could consider challenging or boring for someone with just a hobby. My educational background is applied math with elements of CS, but it's more math than CS. For example, I know very little about functional programming, which I believe can't happen to someone with a purely CS degree.
By mentioning my 400 Project Euler problems, I wasn't in anyway bragging. After all, I'm here asking all these questions, because I feel that I'm lacking. But I didn't want to end up with suggestions of starting with Knuth or CLRS, so I thought the best and quickest way to avoid it was to mention something like the PE.
I realize that I'm being vague in what I want, but I don't know how to be more specific about something that I don't know or I don't know even exists. Basically, anything you would consider a must, but for someone who's not just starting from scratch and who won't be spooked by the more challenging stuff. Anything along the lines of "Prime numbers, computational perspective" or "Computational geometry", perhaps.