InverseErfc
The inverse complementary error function: solves for . Not simply ErfInv(1 - s) — that loses precision exactly in the tail () InverseErfc exists for.
✓ checked against that system · ! checked, with disagreements (hover a chip)Wolfram Language
InverseErfc!InverseErfc(s)solves for .SymPy
erfcinv(s)Rusterfc_inv(s)Typed signatures
| Type | Package | Applies |
|---|---|---|
(number) -> number | analytic | always |
Details
- Built on compute-engine's native
ErfInv, but is computed as a compute-engine expression (exact rational/bignum arithmetic) rather than a plain double, so the finalN()sees a fully precise argument even at — whereErfInvof a literal double would already have dropped several digits. - For , uses to reduce to the case: .
- holds exactly, symbolic included — compute-engine has no general inverse-composition simplification (checked:
ErfInv(Erf(x))stays unevaluated too), soErfcis extended in place with this one structural cancellation.
Plot
Examples
2 more kept as data, checked against the oraclesApplications #
times this, , is the two-sided 99% normal critical value
live ↗InOut
Known value · Wolfram 15.0.0 InverseErfc
About
| Examples | 18 |
|---|---|
| Checked against | Wolfram Language 17 agree1 differSymPy 7 agree7 unevaluated1 errorRust 8 agree3 error |
| Crosswalk | Wolfram LanguageSymPyRust |
Implementation
primitive · kernela kernel — an algorithm over mutable state that would not be readable as a tree
nativeengine
typescriptpackages/symbols/analysis/analytic/src/inverse-erfc.tsmapped
rustpackages/oracle/src/mappings.tsmapped
sympypackages/oracle/src/mappings.ts See also: Erfc