Rounding

Package: FUNCTION_PACKAGE. Registered by createEngine(); for a slimmer engine, register it explicitly (see choosing packages).

Rounding replaces a number with a nearby simpler one: the nearest whole number, a set number of decimal places, or the nearest ten, hundred or thousand. Solve has a form for each, in symbols and in words, so you can say the one you mean.

round on its own goes to the nearest whole number. Give it a place count, or say to N dp, to round to that many decimal places and show exactly that many, trailing zeros kept.

round(3.7) // 4
round(3.14159, 2) // 3.14
3.14159 to 4 dp // 3.1416
1.5 to 2 dp // 1.50
100 to 2 dp // 100.00

The place count is a precision you set on the value, not a global display setting, so a rounded number reads the way you asked and carries that precision into the next line. The rounding is exact where the number has an exact decimal, so a half at the last place goes up rather than down.

1.005 to 2 dp // 1.01
round(2.675, 2) // 2.68

Rounding to a magnitude reads the way it is said. rounded with no target is the nearest whole; up and down force the direction; to nearest <n> rounds to a multiple, and the round magnitude words (ten, hundred, thousand, …) stand in for the number.

5.5 rounded // 6
5.4 rounded up // 6
5.6 rounded down // 5
37 to nearest 10 // 40
490 rounded to nearest hundred // 500
21 rounded up to nearest 5 // 25