PERCENTRANK.EXC in Collabora Office: Exclusive Ranks and Endpoint Comparisons

• 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.EXC(data;value;[significance])

The first two arguments are required. Significance is optional and described in terms of rounded significant digits, but the dedicated section does not state its omitted numerical default. These formulas explicitly request 6.

For distinct ordered observations, the independent exclusive-position model is position/(n+1). The complete source states the exclusive scale without printing that equation. All main examples use exact terminating fractions to keep unverified precision details out of the result.

Enter 5,10,20,30,50,80,130 in A2:A8. With seven observations, the model denominator is eight:

The smallest value is not assigned zero, and the largest is not assigned one. Percentage formatting displays the endpoints here as 12.5% and 87.5%.

For the same seven observations, the inclusive distinct-value model assigns 0 to 5, 0.5 to 30, and 1 to 130.

The exclusive model above assigns 0.125,0.5,0.875 to those three values. Both give the middle observation rank one half, but their endpoint conventions differ.

Use PERCENTRANK.INC when documenting that inclusive question. A formula rename can change a report’s thresholds even when the numeric benchmark is unchanged.

Append 210 and 340 in A9:A10. There are now nine observations, so the exclusive model denominator is ten:

=PERCENTRANK.EXC(A2:A10;30;6)
=PERCENTRANK.EXC(A2:A10;130;6)

Independent expected results: 0.4 and 0.7. Their positions remain fourth and seventh, but the larger comparison set changes their relative positions.

The new maximum 340 has modeled exclusive rank 0.9, still below one. Do not describe an exclusive rank as a ratio to the largest observation.

Return to the original seven-value range. Its highest modeled exclusive rank is 0.875. A rule requiring rank at least 0.9 therefore selects zero observations.

Under the inclusive model, the maximum has rank 1 and would qualify. Neither result is automatically a defect: the selected function defines a different scale. State the convention and threshold together.

A strict comparison above the exclusive median rank 0.5 selects 50,80,130, three of seven observations. That is approximately 42.857143%, not an exact half of a finite odd-sized dataset.

Double the original values to obtain 10,20,40,60,100,160,260. The transformed target 60 retains modeled exclusive rank 0.5, while 260 retains 0.875. Positive scaling changes units but preserves the seven ordered positions.

Keeping the benchmark fixed with an absolute reference is useful when evaluating a list of observed targets, for example =PERCENTRANK.EXC($A$2:$A$8;C2;6).

The source explicitly describes an error for targets outside the total data range. Against the original list, target 200 reaches that condition; its exact error code is unverified. Unobserved interior targets, ties, singletons, mixed types and precision edge cases also remain unresolved.

The namespace COM.MICROSOFT.PERCENTRANK.EXC is documented outside the cited OpenFormula 1.3 standard. This is not a native file-interchange guarantee. PERCENTILE.EXC asks for a value at a chosen exclusive percentile and has separately documented interpolation limits.

• Collabora Office 26.04: PERCENTRANK.EXC.
• 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.