úÎ!"î9      !"#$%&'()*+,-./012345678(c) Edward Kmett 2013-2015BSD3Edward Kmett <ekmett@gmail.com> experimental non-portable Trustworthy  27@AHQV¯ log-domainLog-domain Float and Double values.9 log-domainNegative infinity log-domainIEfficiently and accurately compute the sum of a set of log-domain numbersWhile folding with (+)7 accomplishes the same end, it requires an additional n-2 logarithms to sum nP terms. In addition, here we introduce fewer opportunities for round-off error.;While for small quantities the naive sum accumulates error,:let xs = Prelude.replicate 40000 (Exp 1e-4) :: [Log Float]Prelude.sum xs ~= 4.00e4True&This sum gives a more accurate result,Numeric.Log.sum xs ~= 4.00e4TrueNB:, This does require two passes over the data.(c) Edward Kmett 2013-2015BSD3Edward Kmett <ekmett@gmail.com> experimental non-portableSafe 27@AHVXÊ* log-domainLog-domain Float and Double values, with a sign bit.*+,-*+,-:      !"#$%&'()*+,-./0123456789:;<=log-domain-0.13-inplace Numeric.LogNumeric.Log.SignedLogExplnsum $fFloatingLog $fMonoidLog $fRealLog$fVectorVectorLog$fMVectorMVectorLog $fUnboxLog $fRealFracLog$fFractionalLog$fNumLog $fEnumLog $fMonadLog $fBindLog $fApplyLog$fComonadApplyLog$fApplicativeLog $fComonadLog $fExtendLog$fDistributiveLog$fTraversable1Log$fTraversableLog$fFoldable1Log $fFoldableLog $fNFDataLog $fStorableLog$fHashable1Log $fHashableLog $fFunctorLog $fSerial1Log $fSerialLog$fSerializeLog $fBinaryLog $fReadLog $fShowLog$fSemigroupLog$fEqLog$fOrdLog $fDataLog $fGenericLog SignedLogSLExpsignSLlnSL$fRealFracSignedLog$fFloatingSignedLog$fRealSignedLog$fFractionalSignedLog$fNumSignedLog$fReadSignedLog$fShowSignedLog$fOrdSignedLog $fEqSignedLog$fDataSignedLog$fGenericSignedLognegInf