Ask HN: Math books you recommend just for the way they are written?

Ah, that would've been one of the books I recommended. Shame that he passed with the book in a rather unfinished state.

> Does any book come to your mind along these lines? Books that stop being pedantic where needed to first convey the topic to the reader.

Mathematics books by physicists tend to be just that.

Steven Strogatz's Sync and Infinite Powers

The Foundations of Statistical Inference by L.J. Savage

Winning Ways For Your Mathematical Plays by Berlekamp, Conway and Guy

Not Math books, but David Griffiths Electrodynamics and Quantum Mechanics are such a pleasure to read for their style of writing.

"Calculus Made Easy" by Silvanus P. Thompson, now in public domain.

https://calculusmadeeasy.org/

The book is worth a peek just for this one chapter: https://calculusmadeeasy.org/1.html

"An imaginary tale, the story of sqrt(-1)" -- it's a great romp through both the history of the imaginary unit and actually gently explains a fair bit of complex analysis to the reader, up to contour integration I think

Also, James Gleick's Chaos is a classic, as is the big original beast, Gödel Escher Bach by Douglas Hofstadter.

Understanding Analysis by Stephen Abbott and Algebra by Michael Artin.

Yeah, the intersection of "serious maths book" and "great read" is pretty small.

Since you've already got Jaynes: David Mackay's inference book is also a good: http://www.inference.org.uk/itila/ even if not quite pure maths.

Byrne's translation of Euclid, you can find a digital version here: https://www.c82.net/euclid/

I love Atiyah & MacDonald’s “Introduction to Commutative Algebra”. I was always impressed by how it manages to be so extremely terse and yet so crystal clear and easily readable.

Strang's "Introduction to Linear Algebra". This book is a friend!

Garrity's "All the Math You Missed". Brings you up to speed fast, and has great references.

Feynman & Hibbs was a joy. Physics rather than mathematics, but very enjoyable.

If it wasn't because it's a text introducing a somewhat niche approach to quantum mechanics that didn't gain broad traction, this book would totally have been the Kernighan & Ritchie of quantum mechanics.

Spivak's Calculus. Excellent first proof-based math textbook for anyone interesting in self-studying math.

An Introduction to Complex Analysis for Engineers by Michael Alder

https://cdn.preterhuman.net/texts/math/Mathematics%20-%20An%...

Concrete Mathematics, by Graham, Knuth, and Patashnik. Not only is the subject matter interesting, and, AFAICT not presented all together as a coherent body anywhere else, there are literally notes in the margins that the authors say reflect the comments of students who took courses using the book while it was being developed. Even if you don't care about the subject matter per se, the comments are well worth the read.

Similarly, Proofs from THE BOOK, by Aigner and Ziegler presents some interesting subject matter (short, elegant, and instructive proofs) all in one volume that you'd have to comb through large amounts of mathematical literature to encounter otherwise. The results themselves should be mostly familiar to any grad student or advanced undergrad in mathematics, but the proofs are sure to amuse as well as enlighten. The one unfortunate thing about this book is that the last edition was published almost 25 years ago.

Linderholm's Mathematics Made Difficult is a converse: a book that, despite utter lack of need, leaps into being pedantic. Nevertheless, it remains both amusing and instructive.

https://archive.org/details/mathematics-made-difficult

"Récoltes et semailles" (Harvest and plantings) by Alexander Grothendieck.

https://m.youtube.com/c/themathsorcerer for math book reviews

I particularly enjoy Ahlfors' writing in his complex analysis book.

"Numerical Linear Algebra" by Trefethen and Bau.

"Advanced Mathematical Methods for Scientists and Engineers: Asymptotic Methods and Perturbation Theory" by Bender and Orszag.

I'm not sure if they meet your definition of being well-written exactly as you say above, but these are extremely well-written math books.

This isn't a math book, but I was also a big fan of "Vector Quantization and Signal Compression" by Gersho and Gray.

David MacKay: information theory, inference, learning algorithms.

It’s a classic in the field(s). Available as free pdf. Sets an example for how technical books should be written.

David Williams’ Probability with Martingales is a personal favourite. And I’m not the only one.

- Math with Bad Drawings by Ben Orlin

- Infinite Powers and The Joy of X by Steven Strogatz

- Godel, Escher, Bach by Hofstadter

(Avoiding text books as one's mileage might vary, and having fun depends on the readers' levels as well.)