ERFC.PRECISE in ONLYOFFICE Spreadsheet Editor: Syntax, Examples and Troubleshooting

The underlying mathematical expression is

erfc(x) = (2/sqrt(pi)) * integral from x to infinity of exp(-t^2) dt = 1 - erf(x).

This expression matches the vendor’s x-to-infinity complementary integral. The parameter x is the lower integration limit. The documented signature accepts one argument.

Unlike ERFC(x) in the same official catalogue, this explicitly named ERFC.PRECISE(x) is a distinct documented function name. The official pages describe the same type of complementary integral, but do not give enough performance or release data to promise greater accuracy in any particular installed version. Official ERF reference; official ERFC reference; documented counterpart.

=ERFC.PRECISE(x)

x is dimensionless in the underlying mathematical definition. Ordinary numeric literals and references are supported by the official function entry guide; the vendor’s short ERFC.PRECISE reference does not provide a comprehensive text-to-number coercion table, infinity handling, maximum accepted numeric magnitude or exact error mapping. Those edge cases require native testing. In English-language vendor examples, use an initial = and commas when composing functions with more than one argument. This function itself takes only one argument.

1. Open a blank spreadsheet in ONLYOFFICE Spreadsheet Editor. Record whether you use Desktop Editors or a web/server-hosted Docs editor and note the actual build/version before reporting results. These guides have not executed either edition.
2. Put Case in A1, x (numeric) in B1, ERFC.PRECISE formula in C1 and Expected result (mathematical) in D1. Type the x values in the table below into B2:B19 as genuine numeric cells (do not paste the quote characters around values or include unit labels).
3. Enter =ERFC.PRECISE(B2) into C2 and fill it down to C19. Cell references adjust row by row. Put the expected numbers from column D in D2:D19 as static numeric reference values; they are not native outputs.
4. In E1 enter Absolute difference; in E2 enter =ABS(C2-D2) and fill through E19. In F1 enter Pass tolerance; in F2 enter =E2<0.0000001 and fill down. The F-column is a suggested native test protocol; do not assume the values have been observed to pass. For tiny expected tails, use an application-appropriate relative tolerance rather than a wide absolute cutoff.
5. Use the Formula tab’s Show Formulas when auditing the formula text, and its Calculation control or F9 to recalculate; the vendor instructions document those actions.

Expected values were computed from the mathematical definitions with SciPy and independently cross-checked against 65-digit mpmath; values are rounded for display. No ONLYOFFICE application was executed to produce the table. For the x=8 and x=10 rows, displayed zeros or ones can be misleading when the expected tail is extremely small.

=ERFC.PRECISE(0)
=ERFC.PRECISE(0.5)
=ERFC.PRECISE(-1)

The mathematical expected outputs are 1, about 0.479500122187, and about 1.842700792950 respectively. These formulas are directly usable as proposed examples in a compatible English-function installation; their exact displayed precision and error handling have not been native-tested.

Complement identity: for all real x, erfc(x)+erf(x)=1. Put B2=0.5, compare =ERFC.PRECISE(B2)+ERF.PRECISE(B2) with 1 using a tolerance. The two pieces are expected to be 0.479500122187 and 0.520499877813; the numerical identity is a mathematical cross-check, not an observation in the editor.

Reflection check: erfc(-x)=2-erfc(x). Put 1 in H2 and -1 in H3. =ERFC.PRECISE(H3) is expected to be about 1.84270079295. A value above 1 is legitimate: this is not a conventional 0–1 probability without the 0.5 scaling in the normal-tail formula.

Normal upper-tail probability: enter a standard-normal z-score in H2, then put =0.5*ERFC.PRECISE(H2/SQRT(2)) in I2. For H2=1, independent mathematics gives about 0.158655253931; for H2=2, about 0.0227501319482. Use the actual spreadsheet results only after native testing.

Related verification formula: =0.5*ERFC.PRECISE(H2/SQRT(2)). This single-value expression combines the documented function with arithmetic and SQRT; do not confuse its mathematical expected result with a vendor screenshot or a guaranteed implementation across versions.

For large positive x, 1-ERF.PRECISE(x) can lose significant digits due to subtracting nearly equal values. Mathematically erfc(8)≈1.1224297173E-29 and erfc(10)≈2.0884875838E-45. Whether ONLYOFFICE resolves these tiny tails accurately remains untested, so do not claim numerical precision merely from the function name.

Limits: The function help page does not explicitly establish coercion of logicals or date serials, extreme-value overflow/underflow thresholds, significant digits, treatment of infinities, or round-trip precision. Do not assume compatibility with Excel or another spreadsheet establishes ONLYOFFICE’s runtime behavior.

The currently accessible English ONLYOFFICE Docs Spreadsheet Editor catalog lists ERFC.PRECISE under Engineering, and the specific function page gives its signature. The Insert Function instructions explain manual formulas, function search, and F9 recalculation. Advanced Settings provides language/region controls. These URLs were reviewed on 2026-10-13 UTC.

Not established by this research: first-supported Docs release, first-supported Desktop Editors release, older versions, mobile editions, format-import/export behavior, localized names, exact separator substitution, native outputs or error codes. Separate vendor Docs release history and Desktop Editors history exist, but their mere existence does not prove support in each release. Web Docs and locally installed Desktop Editors must be tested independently. No browser, Desktop installation, native worksheet, screenshots or benchmark timings were used for this guide.

The precise family consists of ERF.PRECISE and ERFC.PRECISE. Compare ERF, which documents an optional upper integration bound, and ERFC, which is a complementary single-input function. Related mathematical tools include square roots and Gaussian cumulative probability transformations. The links lead to checked vendor documentation, not speculative live URLs on Spreadsheet-Tutorials.