FLOOR.PRECISE Function (LibreOffice Calc)

FLOOR.PRECISE moves a number down toward negative infinity to a multiple of a chosen step. The sign of the step is ignored. For positive values this typically reduces the number; for negative values it becomes more negative, not closer to zero. For example, =FLOOR.PRECISE(-45.67;2) returns -46. LibreOffice Help: FLOOR.PRECISE, accessed 2026-10-13.

In one line: =FLOOR.PRECISE(3.45;2) returns 2. This is useful for counting complete groups, rounding budget caps down to step sizes, and other lower-bound calculations. It does not round to the nearest multiple.

=FLOOR.PRECISE(Number [; Significance])

The brackets indicate an optional argument; omit the brackets when entering a formula. These examples use an English-language Calc setup with ; separators and decimal-point numbers. Depending on locale and formula settings, the accepted separator or displayed decimal symbol can differ. Official Help, syntax and default, checked 2026-10-13.

For a nonzero step s, the independent mathematical model is:

s = ABS(Significance)
FLOOR.PRECISE(x; s) ≈ s × mathematical_floor(x / s)

The mathematical floor is the largest integer less than or equal to its input. The symbol ≈ is deliberate: binary floating-point arithmetic in spreadsheets can cause apparent exact boundaries to be evaluated slightly differently. The built-in FLOOR function has additional sign and mode rules; do not substitute it in this expression without separately checking its behavior. For a zero significance, Calc 25.2.3.2 returned 0 and this nonzero-step mathematical model does not apply.

Enter the following formulas into a blank sheet. All results were natively executed on Calc 25.2.3.2 (Linux, 2026-10-13) and independently calculated with Python Decimal floor arithmetic.

Worked example: how many complete kits can you make?

A warehouse has 125 parts. Every assembled kit requires 24 parts, and you cannot sell an incomplete kit.

Calc returns 120 usable parts, 5 complete kits, and 5 leftovers. The same formula can support cartons, manufacturing batches and packaging quantities when both inputs are nonnegative and the unit step is positive.

Protect real-world worksheets from missing, nonnumeric, or negative pack sizes:

=IF(OR(NOT(ISNUMBER(B1));NOT(ISNUMBER(B2));B1<0;B2<=0);"Check stock/kit size";FLOOR.PRECISE(B1;B2))

If B2 represents an indivisible kit size, additionally test B2=INT(B2); this validation is not imposed by the function itself.

Additional practical examples

Budget cap at multiples of five: If D2 contains 87.42, =FLOOR.PRECISE(D2;5) yields 85. This can provide a conservative spending threshold. It is an arithmetic example, not financial advice; the organization determines its actual rounding policy.

Prices down to a five-cent increment: =FLOOR.PRECISE(72.41;0.05) gives 72.40 in the native test. Avoid assuming that this represents a legally required tax, billing, or cash rounding method.

Previous 15-minute boundary: A time serial in A2 can be rounded down using =FLOOR.PRECISE(A2;15/1440). Apply a time/date format to the output. Check date, timezone and daylight-saving requirements independently; the function manipulates numeric day serials, not civil-time rules.

Remainder check: If B3 is the allocated parts and B2 is the strictly positive integer kit size, the remainder is =B1-B3; for the worked example that is 5. An alternative direct count is =INT(B1/B2) for nonnegative inputs; this is equivalent in this specific situation, not a replacement for FLOOR.PRECISE in every sign/domain case.

Compare the related rounding functions:

For a negative number like -45.67, FLOOR.PRECISE(-45.67;2)=-46, while ISO.CEILING(-45.67;2)=-44. The difference is a change of direction, not merely formatting. LibreOffice Help: FLOOR and CEILING families, checked 2026-10-13.

The table contains version-specific native observations from Calc 25.2.3.2 on Linux. It should not be read as a cross-platform guarantee.

Negative values look too negative? This function moves toward negative infinity, so -1.2 with a step of 1 becomes -2. If you need a value closer to zero, consider ISO.CEILING for negative numbers or an explicit conditional formula appropriate to your rules.

A price unexpectedly becomes zero? Check whether the significance cell is zero, or whether a small positive value is below the chosen step. A zero result can be mathematically valid for positive values below the step and can also arise from invalid input when the step itself is zero; these require different handling.

Floating-point precision: In some circumstances, the exact mathematical multiple is not exactly representable in binary. Do not build approval checks around strict equality of arbitrary decimal fractions. Format cell outputs for display, but audit underlying precision separately when policy depends on exact boundary decisions.

Formula treated as text: Confirm the cell’s format is not Text and formulas begin with =. If your locale expects comma decimals, adapt decimal literals and argument separators to the local configuration.

Official Help says FLOOR.PRECISE is available in LibreOffice 4.3 and later. The Help associates it with the extension namespace COM.MICROSOFT.FLOOR.PRECISE and explicitly notes that the function is not in the ODF 1.3 OpenFormula function set. Do not confuse extension namespace support with an inability to save formulas in ODS. LibreOffice Help: FLOOR.PRECISE technical information.

Native execution confirmed: LibreOffice Calc 25.2.3.2 on Linux, 2026-10-13. The online documentation checked on that date carried the LibreOffice 26.8 Help label, so only the testing section makes version-specific native claims. Behavior on Windows/macOS and in other spreadsheet products has not been independently executed. Test .xlsx import/export, negative numbers and zero steps before relying on another program’s calculation.

The evidence workbook executed 35 native numerical assertions for FLOOR.PRECISE. All 35 matched independent Python Decimal floor expectations within the stated numerical tolerance; 12 additional diagnostic formulas were recorded with actual Calc display/error results, not converted into invented passes.

Dated sources and what they support (checked 2026-10-13):

1. LibreOffice Help: FLOOR functions — FLOOR.PRECISE syntax, default, sign behavior, technical namespace and availability.
2. LibreOffice Help: CEILING functions — distinguishing upward-rounding ISO.CEILING and other rounding variants.