Top-down approach:
Instead of building up from the axioms to reach a fact, first, a fact is presented, then it is shown how to use it in practice with examples. Then it is broken down to reveal how and where it comes from. Maybe a proof or something of that sort.
I think “Who Is Fourier? A Mathematical Adventure”, would be an example of the kind of book you are looking for.<p>Its a top-down intro to Fourier transforms that explains the derivation of the transform and why there are complex numbers involved, etc. Starts out very basic and top level, but covers all the important little details by the end.
Numerical Recipes in C: The Art of Scientific Computing<p><a href="https://www.cec.uchile.cl/cinetica/pcordero/MC_libros/NumericalRecipesinC.pdf" rel="nofollow">https://www.cec.uchile.cl/cinetica/pcordero/MC_libros/Numeri...</a>
Do you have a particular CS or math topic that you want to study? Is there a book that you tried, that was too bottom-up?<p>If it's in an applied area, like calculus or statistics, there are usually plenty of books that skip the theory and simply give you recipes to use. If it's a more theoretical area then the theory is presumably why you came, so of course they aren't going to skip it. It's also common to study application-style first, then theory later. So first you take "calculus", then "intro to analysis" which is the same topics as calculus but with the proofs done carefully, and then a more theoretical and abstract approach to analysis.<p>If you start with an abstract textbook, you'll see a bunch of definitions and axioms at the beginning that might make the book feel self-contained like you can start from nothing, but in practice you are supposed to know the basics of the topic before you start such a book. Otherwise you will have a hard time. That might be the difficulty you've encountered.