@@ -131,7 +131,7 @@ The local loss function values then serve as a criterion for choosing the next p
This means that upon adding new data points, only the intervals near the new point needs to have their loss value updated.
The amortized complexity of the point suggestion algorithm is, therefore, $\mathcal{O}(1)$.
The algorithm can be summarized as follows, where `f` is the function to evaluate, and `loss` is the loss function.
The algorithm can be summarized as follows, where `f` is the function to evaluate, `loss` is the loss function, and `heap_push`, `head_pop` and `heap_max` are functions for manipulating a max-heap.