PERCENTRANK in Collabora Office: Observed Percentage Ranks and the Documented Precision Default

• Introduced: original Collabora release and launch date are not established by the reviewed sources.
• Availability: currently documented in Collabora Office 26.04 desktop help.
• Platform scope: online and mobile deployments are not inferred from this desktop-help release.
• Checked: October 14, 2026. Function-specific documentation and independent calculations; no native Collabora execution.

=PERCENTRANK(data;value;[significance])

Data and value are required. Significance is described as the number of significant digits used for the returned percentage value. The source uses rounding terminology but does not provide a complete boundary-sensitive precision algorithm.

The examples use distinct observed numeric values whose ranks terminate exactly within three decimal digits. They avoid inferring nonpositive-significance behavior, a truncation rule or an interior missing-value rule from another spreadsheet.

For the numeric controls, the independent inclusive rank model is (position-1)/(n-1) after sorting. The compact reference does not print that equation; it provides the statistical purpose and the stated range/default rules.

Enter 5,15,30,60,100 in A2:A6:

The position of 60 is fourth among five values, yielding 3/4. A Percentage format displays 0.75 as 75%. It does not require multiplying the stored result by 100 first.

These controls retain exact observed values. The dataset’s unequal numeric gaps do not change the rank assigned to each of its five positions.

=PERCENTRANK(A2:A6;60)
=PERCENTRANK(A2:A6;60;3)
=PERCENTRANK(A2:A6;60;6)

All have independently expected 0.75 for this fixture. The first two compare omitted significance with the explicitly documented default. The third requests more detail than this exact terminating result needs.

This does not establish how a nonterminating rank is rounded at a particular boundary. It also does not authorize a negative or fractional significance argument.

Append 200 in A7 and expand the benchmark to A2:A7. The observed value 60 stays fourth, but the model denominator changes from four to five. Its expected percentage rank becomes 0.6.

The value has not changed. Its relative position changed because the benchmark population changed. Keep the range and reporting period visible beside any score based on percentage rank.

A formula such as =PERCENTRANK($A$2:$A$6;C2) keeps the original comparison range fixed while different observed targets are evaluated in C2.

In the original data, 60 divided by the maximum 100 is 0.6. Its modeled percentage rank is 0.75. One compares magnitudes; the other compares ordered positions.

Add 7 to every original observation, giving 12,22,37,67,107. The transformed target 67 keeps modeled rank 0.75, though its ratio to the maximum is now 67/107, approximately 0.626168.

A positive unit conversion preserves ordering. It need not preserve a value divided by the maximum when the origin changes.

Adding text Pending and a genuinely empty cell to the original range leaves its five ranked numeric observations unchanged under the documented range-ignore rule. A target 200 evaluated against the original range lies above its maximum and reaches the documented error condition; the exact code remains unverified.

The reference does not explicitly settle unobserved interior targets, duplicate ranks, singleton data, formula-empty strings, Boolean values or mixed errors. Those cases are held rather than assigned another application’s behavior.

PERCENTRANK.INC and PERCENTRANK.EXC have separate current sections. Their optional significance syntax does not by itself establish the same default value as plain PERCENTRANK, and none is labeled retired here without lifecycle evidence.

• Collabora Office 26.04: PERCENTRANK.
• Editorial testing standards.

Complete function-specific documentation reviewed. Worked outputs are independent mathematical models or explicitly documented rules, not application screenshots or native test results. Precision, type and error boundaries retain their stated limits.