MINA

What it does

MINA(value1, [value2], ...)

MINA(value1, [value2], …) returns the smallest value in a set, where logical values and text are converted to numbers rather than skipped.

• "Arguments can be the following: numbers; names, arrays, or references that contain numbers; text representations of numbers; or logical values, such as TRUE and FALSE, in a reference."
• "Arguments that contain TRUE evaluate as 1; arguments that contain text or FALSE evaluate as 0 (zero)."
• "If an argument is an array or reference, only values in that array or reference are used. Empty cells and text values in the array or reference are ignored."
• "If the arguments contain no values, MINA returns 0."
• "Arguments that are error values or text that cannot be translated into numbers cause errors."

The consequence that dominates this function. Because text and FALSE convert to 0 rather than being skipped, MINA can never exceed 0 once any text or FALSE is present. And because most spreadsheet data is positive — prices, quantities, durations, distances, counts — that is not an edge case, it is the ordinary case. A column of positive measurements with one "n/a" typed into it has MINA equal to 0, and the 0 appears nowhere in the data.

MINA is therefore not a selector in the sense MIN is: its result need not be one of its inputs. It is the mirror of MAXA's zero clamp, and it is the worse of the two, because the direction of the clamp points into the region where real data lives.

Arguments

Argument boundaries carry no meaning; all slots pool into one reduction.

Try it

Quirks

The coercion table, and a contradiction in it

The two "see below" cells are a documented contradiction, and the Handbook publishes it as a finding rather than choosing a side. Microsoft's MINA page asserts both of the following about text inside a reference:

1. "Arguments that contain text or FALSE evaluate as 0 (zero)."
2. "Empty cells and text values in the array or reference are ignored."

These cannot both hold. If text evaluates as 0 it is counted and drags the minimum down to zero on all-positive data; if it is ignored, the minimum of an all-positive column stays positive. The two readings differ on exactly the case that distinguishes MINA from MIN, and they can differ by the entire magnitude of the data.

The same contradiction appears verbatim on the MAXA page, which is evidence that it is a family-wide editorial defect in the documentation rather than a per-function slip.

What can be said without observing Excel:

• Rule 1 is the one that gives MINA a reason to exist as a separate function. Under rule 2 applied to text, MINA and MIN would differ only on logicals in references.
• Microsoft's own MIN page points at MINA as the way to "include logical values and text representations of numbers in a reference as part of the calculation", which is consistent with rule 1.
• Nothing in the Handbook's record settles it. The probe is one cell wide and is first on the list below.

Result and edge cases

Returns Number. The result need not be one of the inputs, because text and FALSE are converted rather than selected.

• An all-text column. Under rule 1 the answer is 0; under rule 2 it is the no-values case, also 0. The two readings coincide here, so this input does not discriminate between them. The discriminating input is an all-positive numeric column with one text cell.
• TRUE in a reference counts as 1, which is the family's undisputed purpose and the cleanest observable difference from MIN. On data all greater than 1, a stray TRUE becomes the minimum.
• Arrays are consumed by scanning, not lifted elementwise.

Errors

No documented #NUM!, #DIV/0! or empty-input error; the documented empty-input answer is 0.

How to implement

Numerical notes

No floating-point arithmetic beyond the conversions, so no rounding error to analyse. Three observations are worth recording.

The zero clamp is directional, and its direction is the bad one. MAXA's clamp only bites on all-negative data; MINA's bites on all-positive data, which is the common shape. Combined with the no-values rule, MINA reaches 0 by two independent routes, so a hierarchical MINA-of-MINAs over positive data reports 0 if any column is empty or contains any text. An implementer must track "have I seen an admissible value" as a flag separate from the running best and must not initialise the accumulator to 0.

Rust code

aggregate(MINA, "MINA", VALUES, mina)

/// MINA: MIN, with held text read as 0 and held logicals as 1 and 0.
fn mina(
    arguments: &[CalcValue],
    services: &mut EvalServices<'_>,
) -> Result<CalcValue, FunctionFault> {
    extreme(MINA, Held::Values, |n, best| n < best, arguments, services)
}

No implementation has been published with a Handbook test suite for this function.

Version differences

The recorded environments below distinguish a reading from a same-input comparison. Counts describe captured inputs in each environment; overlapping captures and comparisons are not totals of unique worksheet calls.

The record contains no same-input comparison between these environments. Reading different inputs on different builds does not establish either a version change or cross-build agreement.

Recorded environments and comparison details.

Alternatives

Relationships

• MIN — the numbers-only sibling, whose own documentation names MINA as the remedy for wanting logicals and numeric text counted, and whose page carries the same empty-set divergence finding. MINA's page returns the favour: "To exclude logical values and text representations from calculations, use the MIN function instead."
• MAXA — the mirror, carrying the identical documented contradiction. The two pages together are the evidence that the defect is editorial and family-wide.
• The rest of the A suffix family: AVERAGEA, COUNTA, STDEVA, STDEVPA, VARA, VARPA. Shared intent, not one shared rule. COUNTA in particular counts non-empty values without converting them. Read each page.
• SMALL — the order-statistic selector, numbers only, and the surface to reach for when "no data" must not silently become 0.
• Confused with: MIN. A workbook switched from MIN to MINA to "handle the text cells" has changed the answer to 0 on every positive column containing text.

Named variants

No named variants are recorded yet.