Nesting Math.min and Math.max is the shortest clamp: max(value, min) lifts values below the floor, then min(...) against the ceiling caps anything above. The function is total over finite numbers and returns the value unchanged when it is already inside the range. NaN short-circuits both Math.min and Math.max to NaN, which is usually what you want (callers can detect and reject). One gotcha: if min > max, the function still runs but returns min, which silently masks a caller bug; the next accordion guards against that.

Swapped bounds (min accidentally larger than max) produce silent wrong answers in the naive form because min(max(v, 10), 0) collapses to 0. Computing Math.min(a, b) and Math.max(a, b) first normalises the order so callers cannot misuse the helper. Adding an explicit NaN short-circuit is optional but documents intent clearly. Infinity bounds work as expected because Math.min/Math.max treat them transparently, which is handy for half-open ranges (clamp from 0 to Infinity).

Many UI features (slider step, pagination index, pixel snapping) need an integer in the range, so rounding after clamp is the standard recipe. Switching the rounding mode lets callers pick round (default), floor (truncate toward zero for positives), or ceil (snap up), which covers nearly every real use case. Doing the clamp before the round avoids drifting just past the boundary on 0.5 rounding behaviour around the bounds. For half-even or banker's rounding, swap in your own snapper since Math.round always rounds toward positive infinity on .5.