It is the value of W(1) where W is Lambert's W function. The name is derived from the alternate name for Lambert's W function, the omega function. The value of Ω is approximately
0.5671432904097838729999686622... (sequence A030178 in the OEIS).
Fixed point representation
The defining identity can be expressed, for example, as
ln(1Ω)=Ω.
or
−ln(Ω)=Ω
or
e−Ω=Ω.
In Mathematica
Plot[{x, Exp[-x], -Log[x]}, {x, 0, 1}]
Computation
One can calculate Ω iteratively, by starting with an initial guess Ω0, and considering the sequence
Ωn+1=e−Ωn.
init = 0.5;
FixedPoint[Exp[-#] &, init]
This sequence will converge towards Ω as n→∞. This convergence is because Ω is an attractive fixed point of the function e−x.
It is much more efficient to use the iteration
Ωn+1=1+Ωn1+eΩn,
because the function
f(x)=1+x1+ex=1+x1+e−x⋅e−x
has the same fixed point but features a zero derivative at this fixed point, therefore the convergence is quadratic (the number of correct digits is roughly doubled with each iteration).
Integral representation
A beautiful identity due to Victor Adamchik[1] is given by the relationship
∫−∞∞dt(et−t)2+π2=11+Ω
or
Ω=1∫−∞∞dt(et−t)2+π2−1.
Irrationality and transcendence
Ω can be proven irrational from the fact that e is transcendental; if Ω were rational, then there would exist integers p and q such that
pq=Ω
so that
1=pe(pq)q
and
e=(qp)(qp)=qqpqp
The number e would therefore be algebraic of degree p. However e is transcendental, so Ω must be irrational.
Ω is in fact transcendental as the direct consequence of Lindemann–Weierstrass theorem. If Ω were algebraic, e−Ω would be transcendental; but Ω=exp(-Ω), so these cannot both be true.