GEOMDIST Function (Gnumeric)

• Introduced: First release not established from the evidence cited here.
• Supported in: Gnumeric 1.12.62, confirmed by its function registration. No retirement is established.
• Scope: Desktop Gnumeric with the Statistics Functions plugin. Tables use English comma-separated formulas.
• Checked: October 14, 2026. Tables were recalculated in 1.12.57 and finite reference values independently checked. Version 1.12.62 was source-checked, not executed.

GEOMDIST counts failures before the first success in independent trials with constant success probability p. For k≥0, the mass is p(1−p)^k and the cumulative probability is 1−(1−p)^(k+1). Counting the trial on which success occurs gives k+1 instead, so the two common geometric conventions differ by one. Official function reference.

=GEOMDIST(k,p,cumulative)

All arguments shown are required. The version-pinned help and evaluator establish the Gnumeric contract.

The table records the tested release’s outputs, with finite mathematical references rounded to twelve decimal places. Keep full precision when using a result in a later calculation.

Suppose each independent attempt succeeds with probability 0.25. Put 3 in A2 and 0.25 in B2. =GEOMDIST(A2,B2,FALSE()) gives 0.10546875, the probability of three failures followed by success on the fourth attempt. =GEOMDIST(A2,B2,TRUE()) gives 0.68359375, the probability of success by the fourth attempt. An attempt limit should therefore be translated into a failure-count limit before calling the function.

The probability of still having no success after four attempts is =1-GEOMDIST(3,0.25,TRUE()), 0.31640625, which also equals 0.75^4. Success exactly on attempt five requires four preceding failures, giving =GEOMDIST(4,0.25,FALSE())=0.0791015625.

At p=0.25 the expected number of failures is (1−p)/p=3, while the expected number of attempts is 1/p=4. These means are not deadlines or guarantees. The model’s constant-probability assumption fails if attempts learn from one another or the process changes after each failure.

Zero-probability caution: the outer function wrapper permits p=0, but the probability helpers reject it. Both tested modes return #NUM! at p=0, rather than the limiting finite-k probability zero. For ordinary use choose 0<p≤1. The failure count is floored; a negative fraction such as −0.1 becomes invalid rather than zero. The cumulative selector is required. At p=1, success is immediate: zero failures has mass 1. The geometric helpers explain the stricter probability guard.

These additional outcomes were observed in the same native recalculation. F5 was genuinely empty; it is not an omitted argument. Missing inputs still need deliberate validation, even if they are converted to a numeric zero.

The worked calculations do not certify every locale, imported text value, extreme parameter or later binary version. Shared numerical helpers can change even when a function’s wrapper remains the same.

• Gnumeric function reference
• Gnumeric 1.12.62 evaluator

The reference and boundary tables use recorded 1.12.57 command-line recalculation. Further model examples were independently calculated from their stated inputs; they are not additional native test claims.