Andrew He

In the randomized General Matching code, the code only knows how to extract a perfect matching, so we add up to N additional vertices before computing the matrix inverse which...

- [x] div_vector; container bucketed by sqrt - [x] multiplication - [x] inverse / sqrt - [x] prime count (exp/log) - [ ] fraction for euler transform - [ ]...

- [ ] seg_tree - [ ] bit - [ ] fft

Algorithms - [x] multiply_inverser - [x] polynomial / power series class - [x] integ/deriv/exp/log/pow - [ ] interpolation/multipoint - [ ] GCD Interface - [ ] square - [ ]...

See old book

- [x] Point2D - [x] 2d utilities - [ ] 2d hull - [ ] Circumcircle - [ ] min-enclosing disk - [x] Point3D - [ ] 3d hull -...

Maybe we want faster 3d hull?

We should put NT things into the book with fancier Euclidean stuff. For instance, `count(A * x + B * y = 0, y >= 0`. Code for that problem...

LCA

4

More tree primitives, like distance to LCA, last node before LCA, kth node on path, etc