running min and running max are not stable when values are equal
running min and running max are not stable when values are equal: a task in repo2rlenv-swe-next (Harbor dataset). Python's built-in min(x, y) and max(x, y) are stable : when x == y , they return x (the first/oldest value). The running min and running max functions from more-itertools don't respect…
The task
Python's built-in `min(x, y)` and `max(x, y)` are **stable**: when `x == y`, they return `x` (the first/oldest value). The `running_min` and `running_max` functions from `more-itertools` don't respect this — when multiple equal values appear in the sliding window, the *newest* one wins instead of the oldest.
Part of FineEnvs/repo2rlenv-swe-next.