Complex logarithm

In mathematics, a complex logarithm is a generalization of the natural logarithm to nonzero complex numbers. The term refers to one of the following, which are strongly related:

• A complex logarithm of a nonzero complex number z {\displaystyle z} , defined to be any complex number w {\displaystyle w} for which e w = z {\displaystyle e^{w}=z} .^[1][2] Such a number w {\displaystyle w} is denoted by log ⁡ z {\displaystyle \log z} .^[1] If z {\displaystyle z} is given in polar form as z = r e i θ {\displaystyle z=re^{i\theta }} , where r {\displaystyle r} and θ {\displaystyle \theta } are real numbers with r > 0 {\displaystyle r>0} , then ln ⁡ r + i θ {\displaystyle \ln r+i\theta } is one logarithm of z {\displaystyle z} , and all the complex logarithms of z {\displaystyle z} are exactly the numbers of the form ln ⁡ r + i ( θ + 2 π k ) {\displaystyle \ln r+i\left(\theta +2\pi k\right)} for integers k {\displaystyle k} .^[1][2] These logarithms are equally spaced along a vertical line in the complex plane.
• A complex-valued function log : U → C {\displaystyle \log \colon U\to \mathbb {C} } , defined on some subset U {\displaystyle U} of the set C ∗ {\displaystyle \mathbb {C} ^{*}} of nonzero complex numbers, satisfying e log ⁡ z = z {\displaystyle e^{\log z}=z} for all z {\displaystyle z} in U {\displaystyle U} . Such complex logarithm functions are analogous to the real logarithm function ln : R > 0 → R {\displaystyle \ln \colon \mathbb {R} _{>0}\to \mathbb {R} } , which is the inverse of the real exponential function and hence satisfies e^{ln x} = x for all positive real numbers x. Complex logarithm functions can be constructed by explicit formulas involving real-valued functions, by integration of 1 / z {\displaystyle 1/z} , or by the process of analytic continuation.

There is no continuous complex logarithm function defined on all of C ∗ {\displaystyle \mathbb {C} ^{*}} . Ways of dealing with this include branches, the associated Riemann surface, and partial inverses of the complex exponential function. The principal value defines a particular complex logarithm function Log : C ∗ → C {\displaystyle \operatorname {Log} \colon \mathbb {C} ^{*}\to \mathbb {C} } that is continuous except along the negative real axis; on the complex plane with the negative real numbers and 0 removed, it is the analytic continuation of the (real) natural logarithm.

For a function to have an inverse, it must map distinct values to distinct values; that is, it must be injective. But the complex exponential function is not injective, because e w + 2 π i k = e w {\displaystyle e^{w+2\pi ik}=e^{w}} for any complex number w {\displaystyle w} and integer k {\displaystyle k} , since adding i θ {\displaystyle i\theta } to z {\displaystyle z} has the effect of rotating e w {\displaystyle e^{w}} counterclockwise θ {\displaystyle \theta } radians. So the points

… , w − 4 π i , w − 2 π i , w , w + 2 π i , w + 4 π i , … , {\displaystyle \ldots ,\;w-4\pi i,\;w-2\pi i,\;w,\;w+2\pi i,\;w+4\pi i,\;\ldots ,}

equally spaced along a vertical line, are all mapped to the same number by the exponential function. This means that the exponential function does not have an inverse function in the standard sense.^[3][4] There are two solutions to this problem.

One is to restrict the domain of the exponential function to a region that does not contain any two numbers differing by an integer multiple of 2 π i {\displaystyle 2{\mathit {\pi i}}} : this leads naturally to the definition of branches of log ⁡ z {\displaystyle \log z} , which are certain functions that single out one logarithm of each number in their domains. This is analogous to the definition of arcsin ⁡ x {\displaystyle \arcsin x} on [ − 1 , 1 ] {\displaystyle [-1,1]} as the inverse of the restriction of sin ⁡ θ {\displaystyle \sin \theta } to the interval [ − π / 2 , π / 2 ] {\displaystyle [-\pi /2,\pi /2]} : there are infinitely many real numbers θ {\displaystyle \theta } with sin ⁡ θ = x {\displaystyle \sin \theta =x} , but one arbitrarily chooses the one in [ − π / 2 , π / 2 ] {\displaystyle [-\pi /2,\pi /2]} .

Another way to resolve the indeterminacy is to view the logarithm as a function whose domain is not a region in the complex plane, but a Riemann surface that covers the punctured complex plane in an infinite-to-1 way.

Branches have the advantage that they can be evaluated at complex numbers. On the other hand, the function on the Riemann surface is elegant in that it packages together all branches of the logarithm and does not require an arbitrary choice as part of its definition.

For each nonzero complex number z {\displaystyle z} , the principal value Log ⁡ z {\displaystyle \operatorname {Log} z} is the logarithm whose imaginary part lies in the interval ( − π , π ] {\displaystyle (-\pi ,\pi ]} .^[2] The expression Log ⁡ 0 {\displaystyle \operatorname {Log} 0} is left undefined since there is no complex number w {\displaystyle w} satisfying e w = 0 {\displaystyle e^{w}=0} .^[1]

When the notation log ⁡ z {\displaystyle \log z} appears without any particular logarithm having been specified, it is generally best to assume that the principal value is intended. In particular, this gives a value consistent with the real value of ln ⁡ z {\displaystyle \ln z} when z {\displaystyle z} is a positive real number. The capitalization in the notation Log {\displaystyle {\text{Log}}} is used by some authors^[2] to distinguish the principal value from other logarithms of z . {\displaystyle z.}

The polar form of a nonzero complex number z = x + y i {\displaystyle z=x+yi} is z = r e i θ {\displaystyle z=re^{i\theta }} , where r = | z | = x 2 + y 2 {\textstyle r=|z|={\sqrt {x^{2}+y^{2}}}} is the absolute value of z {\displaystyle z} , and θ {\displaystyle \theta } is its argument. The absolute value is real and positive. The argument is defined up to addition of an integer multiple of 2π. Its principal value is the value that belongs to the interval ( − π , π ] {\displaystyle (-\pi ,\pi ]} , which is expressed as atan2 ⁡ ( y , x ) {\displaystyle \operatorname {atan2} (y,x)} .

This leads to the following formula for the principal value of the complex logarithm:

Log ⁡ z = ln ⁡ r + i θ = ln ⁡ | z | + i Arg ⁡ z = ln ⁡ x 2 + y 2 + i atan2 ⁡ ( y , x ) . {\displaystyle \operatorname {Log} z=\ln r+i\theta =\ln |z|+i\operatorname {Arg} z=\ln {\sqrt {x^{2}+y^{2}}}+i\operatorname {atan2} (y,x).}

For example, Log ⁡ ( − 3 i ) = ln ⁡ 3 − π i / 2 {\displaystyle \operatorname {Log} (-3i)=\ln 3-\pi i/2} , and Log ⁡ ( − 3 ) = ln ⁡ 3 + π i {\displaystyle \operatorname {Log} (-3)=\ln 3+\pi i} .

Another way to describe Log ⁡ z {\displaystyle \operatorname {Log} z} is as the inverse of a restriction of the complex exponential function, as in the previous section. The horizontal strip S {\displaystyle S} consisting of complex numbers w = x + y i {\displaystyle w=x+yi} such that − π < y ≤ π {\displaystyle -\pi <y\leq \pi } is an example of a region not containing any two numbers differing by an integer multiple of 2 π i {\displaystyle 2\pi i} , so the restriction of the exponential function to S {\displaystyle S} has an inverse. In fact, the exponential function maps S {\displaystyle S} bijectively to the punctured complex plane C ∗ = C ∖ { 0 } {\displaystyle \mathbb {C} ^{*}=\mathbb {C} \setminus \{0\}} , and the inverse of this restriction is Log : C ∗ → S {\displaystyle \operatorname {Log} \colon \mathbb {C} ^{*}\to S} . The conformal mapping section below explains the geometric properties of this map in more detail.

On the region C − R ≤ 0 {\displaystyle \mathbb {C} -\mathbb {R} _{\leq 0}} consisting of complex numbers that are not negative real numbers or 0, the function Log ⁡ z {\displaystyle \operatorname {Log} z} is the analytic continuation of the natural logarithm. The values on the negative real line can be obtained as limits of values at nearby complex numbers with positive imaginary parts.

Not all identities satisfied by ln {\displaystyle \ln } extend to complex numbers. It is true that e Log ⁡ z = z {\displaystyle e^{\operatorname {Log} z}=z} for all z ≠ 0 {\displaystyle z\neq 0} (this is what it means for Log ⁡ z {\displaystyle \operatorname {Log} z} to be a logarithm of z {\displaystyle z} ), but the identity Log ⁡ ( e z ) = z {\displaystyle \operatorname {Log} (e^{z})=z} fails for z {\displaystyle z} outside the strip S {\displaystyle S} . For this reason, one cannot always apply Log {\displaystyle {\text{Log}}} to both sides of an identity e z = e w {\displaystyle e^{z}=e^{w}} to deduce z = w {\displaystyle z=w} . Also, the identity Log ⁡ ( z 1 z 2 ) = Log ⁡ z 1 + Log ⁡ z 2 {\displaystyle \operatorname {Log} (z_{1}z_{2})=\operatorname {Log} z_{1}+\operatorname {Log} z_{2}} can fail: the two sides can differ by an integer multiple of 2 π i {\displaystyle 2\pi i} ;^[1] for instance,

Log ⁡ ( ( − 1 ) i ) = Log ⁡ ( − i ) = ln ⁡ ( 1 ) − π i 2 = − π i 2 , {\displaystyle \operatorname {Log} ((-1)i)=\operatorname {Log} (-i)=\ln(1)-{\frac {\pi i}{2}}=-{\frac {\pi i}{2}},}

but

Log ⁡ ( − 1 ) + Log ⁡ ( i ) = ( ln ⁡ ( 1 ) + π i ) + ( ln ⁡ ( 1 ) + π i 2 ) = 3 π i 2 ≠ − π i 2 . {\displaystyle \operatorname {Log} (-1)+\operatorname {Log} (i)=\left(\ln(1)+\pi i\right)+\left(\ln(1)+{\frac {\pi i}{2}}\right)={\frac {3\pi i}{2}}\neq -{\frac {\pi i}{2}}.}

The function Log ⁡ z {\displaystyle \operatorname {Log} z} is discontinuous at each negative real number, but continuous everywhere else in C ∗ {\displaystyle \mathbb {C} ^{*}} . To explain the discontinuity, consider what happens to arg ⁡ z {\displaystyle \arg z} as z {\displaystyle z} approaches a negative real number a {\displaystyle a} . If z {\displaystyle z} approaches a {\displaystyle a} from above, then arg ⁡ z {\displaystyle \arg z} approaches π , {\displaystyle \pi ,} which is also the value of arg ⁡ a {\displaystyle \arg a} itself. But if z {\displaystyle z} approaches a {\displaystyle a} from below, then arg ⁡ z {\displaystyle \arg z} approaches − π . {\displaystyle -\pi .} So arg ⁡ z {\displaystyle \arg z} "jumps" by 2 π {\displaystyle 2\pi } as z {\displaystyle z} crosses the negative real axis, and similarly Log ⁡ z {\displaystyle \operatorname {Log} z} jumps by 2 π i . {\displaystyle 2\pi i.}

Is there a different way to choose a logarithm of each nonzero complex number so as to make a function L ⁡ ( z ) {\displaystyle \operatorname {L} (z)} that is continuous on all of C ∗ {\displaystyle \mathbb {C} ^{*}} ? The answer is no. To see why, imagine tracking such a logarithm function along the unit circle, by evaluating L ⁡ ( e i θ ) {\displaystyle \operatorname {L} \left(e^{i\theta }\right)} as θ {\displaystyle \theta } increases from 0 {\displaystyle 0} to 2 π {\displaystyle 2\pi } . If L ⁡ ( z ) {\displaystyle \operatorname {L} (z)} is continuous, then so is L ⁡ ( e i θ ) − i θ {\displaystyle \operatorname {L} \left(e^{i\theta }\right)-i\theta } , but the latter is a difference of two logarithms of e i θ , {\displaystyle e^{i\theta },} so it takes values in the discrete set 2 π i Z , {\displaystyle 2\pi i\mathbb {Z} ,} so it is constant. In particular, L ⁡ ( e 2 π i ) − 2 π i = L ⁡ ( e 0 ) − 0 {\displaystyle \operatorname {L} \left(e^{2\pi i}\right)-2\pi i=\operatorname {L} \left(e^{0}\right)-0} , which contradicts L ⁡ ( e 2 π i ) = L ⁡ ( e 0 ) {\displaystyle \operatorname {L} \left(e^{2\pi i}\right)=\operatorname {L} \left(e^{0}\right)} .

To obtain a continuous logarithm defined on complex numbers, it is hence necessary to restrict the domain to a smaller subset U {\displaystyle U} of the complex plane. Because one of the goals is to be able to differentiate the function, it is reasonable to assume that the function is defined on a neighborhood of each point of its domain; in other words, U {\displaystyle U} should be an open set. Also, it is reasonable to assume that U {\displaystyle U} is connected, since otherwise the function values on different components of U {\displaystyle U} could be unrelated to each other. All this motivates the following definition:

A branch of log ⁡ z {\displaystyle \log z} is a continuous function L ⁡ ( z ) {\displaystyle \operatorname {L} (z)} defined on a connected open subset U {\displaystyle U} of the complex plane such that L ⁡ ( z ) {\displaystyle \operatorname {L} (z)} is a logarithm of z {\displaystyle z} for each z {\displaystyle z} in U {\displaystyle U} .^[2]

For example, the principal value defines a branch on the open set where it is continuous, which is the set C − R ≤ 0 {\displaystyle \mathbb {C} -\mathbb {R} _{\leq 0}} obtained by removing 0 and all negative real numbers from the complex plane.

Another example: The Mercator series

ln ⁡ ( 1 + u ) = ∑ n = 1 ∞ ( − 1 ) n + 1 n u n = u − u 2 2 + u 3 3 − ⋯ {\displaystyle \ln(1+u)=\sum _{n=1}^{\infty }{\frac {(-1)^{n+1}}{n}}u^{n}=u-{\frac {u^{2}}{2}}+{\frac {u^{3}}{3}}-\cdots }

converges locally uniformly for | u | < 1 {\displaystyle |u|<1} , so setting z = 1 + u {\displaystyle z=1+u} defines a branch of log ⁡ z {\displaystyle \log z} on the open disk of radius 1 centered at 1. (Actually, this is just a restriction of Log ⁡ z {\displaystyle \operatorname {Log} z} , as can be shown by differentiating the difference and comparing values at 1.)

Once a branch is fixed, it may be denoted log ⁡ z {\displaystyle \log z} if no confusion can result. Different branches can give different values for the logarithm of a particular complex number, however, so a branch must be fixed in advance (or else the principal branch must be understood) in order for " log ⁡ z {\displaystyle \log z} " to have a precise unambiguous meaning.

In some literature, the notation ln k ⁡ z {\displaystyle \ln _{k}z} is used to explicitly denote the k {\displaystyle k} -th branch of the complex logarithm. This notation is particularly useful when working with multi-valued logarithms in complex analysis and topology. It was first introduced in the paper Unwinding the Branches of the Lambert W function^[5] and was later referenced in the work of David Jeffrey.^[6]

The argument above involving the unit circle generalizes to show that no branch of log ⁡ z {\displaystyle \log z} exists on an open set U {\displaystyle U} containing a closed curve that winds around 0. One says that log ⁡ z {\displaystyle \log z} has a branch point at 0. To avoid containing closed curves winding around 0, U {\displaystyle U} is typically chosen as the complement of a ray or curve in the complex plane going from 0 (inclusive) to infinity in some direction. In this case, the curve is known as a branch cut. For example, the principal branch has a branch cut along the negative real axis.

If the function L ⁡ ( z ) {\displaystyle \operatorname {L} (z)} is extended to be defined at a point of the branch cut, it will necessarily be discontinuous there; at best it will be continuous "on one side", like Log ⁡ z {\displaystyle \operatorname {Log} z} at a negative real number.

Like every holomorphic function, the complex logarithm can be represented locally – near any point in its domain – with a power series; that is, it is also analytic. Because the logarithm is the integral of 1 / z {\displaystyle 1/z} , its power series can be found by term-by-term integration of a power series for z ↦ 1 / z {\displaystyle z\mapsto 1/z} . A power series for z ↦ 1 / z {\displaystyle z\mapsto 1/z} centered at any point z 0 ≠ 0 {\displaystyle z_{0}\neq 0} can itself be found by considering 1 / z {\displaystyle 1/z} as the sum of a geometric series with initial term 1 / z 0 {\displaystyle 1/z_{0}} and common ratio ( z − z 0 ) / ( − z 0 ) {\displaystyle (z-z_{0})/(-z_{0})} .^{[citation needed]} 1 z = 1 z 0 ⋅ 1 1 − z − z 0 − z 0 = ∑ n = 0 ∞ 1 z 0 ( z − z 0 − z 0 ) n = ∑ n = 0 ∞ ( − 1 ) n z 0 n + 1 ( z − z 0 ) n . {\displaystyle {\frac {1}{z}}={\frac {1}{z_{0}}}\cdot {\frac {1}{1-{\dfrac {z-z_{0}}{-z_{0}}}}}=\sum _{n=0}^{\infty }{\frac {1}{z_{0}}}\left({\frac {z-z_{0}}{-z_{0}}}\right)^{n}=\sum _{n=0}^{\infty }{\frac {(-1)^{n}}{z_{0}^{n+1}}}(z-z_{0})^{n}.} This series converges in the disk | z − z 0 | < | z 0 | {\displaystyle |z-z_{0}|<|z_{0}|} , the largest disk not containing the origin (where there is a pole). After term-by-term integration: log ⁡ z = log ⁡ z 0 + ∑ n = 1 ∞ ( − 1 ) n − 1 n z 0 n ( z − z 0 ) n . {\displaystyle \log z=\log z_{0}+\sum _{n=1}^{\infty }{\frac {(-1)^{n-1}}{nz_{0}^{n}}}(z-z_{0})^{n}.}

The representation of 1 z {\displaystyle {\tfrac {1}{z}}} by a power series with the center z 0 ≠ .0 {\displaystyle z_{0}\not =.0} provides a Taylor series representation for a branch of the logarithm on D r ( z 0 ) ¯ {\displaystyle {\overline {D_{r}(z_{0})}}} with r < | z 0 | {\displaystyle r<|z_{0}|} and c = ln ⁡ ( | z 0 | ) + i ⋅ t {\displaystyle c=\ln(|z_{0}|)+i\cdot t} similar to the Mercator series with an arbitrary center z 0 = | z 0 | ⋅ e i t ≠ 0 {\displaystyle z_{0}=|z_{0}|\cdot e^{it}\not =0} :

F ( z ) = c + ∑ n = 0 ∞ ( − 1 ) n z 0 n + 1 ⋅ ( n + 1 ) ⋅ ( z − z 0 ) n + 1 {\displaystyle F(z)=c+\sum _{n=0}^{\infty }{\frac {(-1)^{n}}{z_{0}^{n+1}\cdot (n+1)}}\cdot (z-z_{0})^{n+1}}

The corresponding branch of the logarithm for F {\displaystyle F} is defined on the domain U := C ∖ { z ∈ C : z = λ ⋅ z 0  with  0 ≥ λ ∈ R } {\displaystyle U:=\mathbb {C} \setminus \{z\in \mathbb {C} \,:\,z=\lambda \cdot z_{0}{\mbox{ with }}0\geq \lambda \in \mathbb {R} \}} .

By removing the negative imaginary axis from the domain of the branch of the logarithm

U i := C ∖ { z ∈ C : z = λ ⋅ i  with  0 ≥ λ ∈ R } {\displaystyle U_{i}:=\mathbb {C} \setminus \{z\in \mathbb {C} \,:\,z=\lambda \cdot i{\mbox{ with }}0\geq \lambda \in \mathbb {R} \}}

a branch of the logarithm can be defined for all x ∈ R ∖ { 0 } {\displaystyle x\in \mathbb {R} \setminus \{0\}} with:

ln ⁡ ( − x ) = ln ⁡ ( x ) + i ⋅ π {\displaystyle \ln(-x)=\ln(x)+i\cdot \pi } with x ∈ R {\displaystyle x\in \mathbb {R} } and x < 0 {\displaystyle x<0} .

The branch of the logarithm on U i {\displaystyle U_{i}} provides the standard logarithm ln ⁡ ( x ) {\displaystyle \ln(x)} for x ∈ R {\displaystyle x\in \mathbb {R} } and x > 0 {\displaystyle x>0} .

Each branch L ⁡ ( z ) {\displaystyle \operatorname {L} (z)} of log ⁡ z {\displaystyle \log z} on an open set U {\displaystyle U} is the inverse of a restriction of the exponential function, namely the restriction to the image L ⁡ ( U ) {\displaystyle \operatorname {L} (U)} . Since the exponential function is holomorphic (that is, complex differentiable) with nonvanishing derivative, the complex analogue of the inverse function theorem applies. It shows that L ⁡ ( z ) {\displaystyle \operatorname {L} (z)} is holomorphic on U {\displaystyle U} , and L ′ ⁡ ( z ) = 1 / z {\displaystyle \operatorname {L} '(z)=1/z} for each z {\displaystyle z} in U {\displaystyle U} .^[2] Another way to prove this is to check the Cauchy–Riemann equations in polar coordinates.^[2]

The function ln ⁡ ( x ) {\displaystyle \ln(x)} for real x > 0 {\displaystyle x>0} can be constructed by the formula ln ⁡ ( x ) = ∫ 1 x d u u . {\displaystyle \ln(x)=\int _{1}^{x}{\frac {du}{u}}.} If the range of integration started at a positive number a {\displaystyle a} other than 1, the formula would have to be ln ⁡ ( x ) = ln ⁡ ( a ) + ∫ a x d u u {\displaystyle \ln(x)=\ln(a)+\int _{a}^{x}{\frac {du}{u}}} instead.

In developing the analogue for the complex logarithm, there is an additional complication: the definition of the complex integral requires a choice of path. Fortunately, if the integrand is holomorphic, then the value of the integral is unchanged by deforming the path (while holding the endpoints fixed), and in a simply connected region U {\displaystyle U} (a region with "no holes"), any path from a {\displaystyle a} to z {\displaystyle z} inside U {\displaystyle U} can be continuously deformed inside U {\displaystyle U} into any other. All this leads to the following:

If U {\displaystyle U} is a simply connected open subset of C {\displaystyle \mathbb {C} } not containing 0, then a branch of log ⁡ z {\displaystyle \log z} defined on U {\displaystyle U} can be constructed by choosing a starting point a {\displaystyle a} in U {\displaystyle U} , choosing a logarithm b {\displaystyle b} of a {\displaystyle a} , and defining L ( z ) = b + ∫ a z d w w {\displaystyle L(z)=b+\int _{a}^{z}{\frac {dw}{w}}} for each z {\displaystyle z} in U {\displaystyle U} .^[7]

Any holomorphic map f : U → C {\displaystyle f\colon U\to \mathbb {C} } satisfying f ′ ( z ) ≠ 0 {\displaystyle f'(z)\neq 0} for all z ∈ U {\displaystyle z\in U} is a conformal map, which means that if two curves passing through a point a {\displaystyle a} of U {\displaystyle U} form an angle α {\displaystyle \alpha } (in the sense that the tangent lines to the curves at a {\displaystyle a} form an angle α {\displaystyle \alpha } ), then the images of the two curves form the same angle α {\displaystyle \alpha } at f ( a ) {\displaystyle f(a)} . Since a branch of log ⁡ z {\displaystyle \log z} is holomorphic, and since its derivative 1 / z {\displaystyle 1/z} is never 0, it defines a conformal map.

For example, the principal branch w = Log ⁡ z {\displaystyle w=\operatorname {Log} z} , viewed as a mapping from C − R ≤ 0 {\displaystyle \mathbb {C} -\mathbb {R} _{\leq 0}} to the horizontal strip defined by | Im ⁡ z | < π {\displaystyle \left|\operatorname {Im} z\right|<\pi } , has the following properties, which are direct consequences of the formula in terms of polar form:

• Circles^[8] in the z-plane centered at 0 are mapped to vertical segments in the w-plane connecting a − π i {\displaystyle a-\pi i} to a + π i {\displaystyle a+\pi i} , where a {\displaystyle a} is the real log of the radius of the circle.
• Rays emanating from 0 in the z-plane are mapped to horizontal lines in the w-plane.

Each circle and ray in the z-plane as above meet at a right angle. Their images under Log are a vertical segment and a horizontal line (respectively) in the w-plane, and these too meet at a right angle. This is an illustration of the conformal property of Log.

The various branches of log ⁡ z {\displaystyle \log z} cannot be glued to give a single continuous function log : C ∗ → C {\displaystyle \log \colon \mathbb {C} ^{*}\to \mathbb {C} } because two branches may give different values at a point where both are defined. Compare, for example, the principal branch Log ⁡ z {\displaystyle \operatorname {Log} z} on C − R ≤ 0 {\displaystyle \mathbb {C} -\mathbb {R} _{\leq 0}} with imaginary part θ {\displaystyle \theta } in ( − π , π ) {\displaystyle (-\pi ,\pi )} and the branch L ⁡ ( z ) {\displaystyle \operatorname {L} (z)} on C − R ≥ 0 {\displaystyle \mathbb {C} -\mathbb {R} _{\geq 0}} whose imaginary part θ {\displaystyle \theta } lies in ( 0 , 2 π ) {\displaystyle (0,2\pi )} . These agree on the upper half plane, but not on the lower half plane. So it makes sense to glue the domains of these branches only along the copies of the upper half plane. The resulting glued domain is connected, but it has two copies of the lower half plane. Those two copies can be visualized as two levels of a parking garage, and one can get from the Log {\displaystyle {\text{Log}}} level of the lower half plane up to the L {\displaystyle {\text{L}}} level of the lower half plane by going 2 π {\displaystyle 2\pi } radians counterclockwise around 0, first crossing the positive real axis (of the Log {\displaystyle {\text{Log}}} level) into the shared copy of the upper half plane and then crossing the negative real axis (of the L {\displaystyle {\text{L}}} level) into the L {\displaystyle {\text{L}}} level of the lower half plane.

One can continue by gluing branches with imaginary part θ {\displaystyle \theta } in ( π , 3 π ) {\displaystyle (\pi ,3\pi )} , in ( 2 π , 4 π ) {\displaystyle (2\pi ,4\pi )} , and so on, and in the other direction, branches with imaginary part θ {\displaystyle \theta } in ( − 2 π , 0 ) {\displaystyle (-2\pi ,0)} , in ( − 3 π , − π ) {\displaystyle (-3\pi ,-\pi )} , and so on. The final result is a connected surface that can be viewed as a spiraling parking garage with infinitely many levels extending both upward and downward. This is the Riemann surface R {\displaystyle R} associated to log ⁡ z {\displaystyle \log z} .^[9]

A point on R {\displaystyle R} can be thought of as a pair ( z , θ ) {\displaystyle (z,\theta )} where θ {\displaystyle \theta } is a possible value of the argument of z {\displaystyle z} . In this way, R can be embedded in C × R ≈ R 3 {\displaystyle \mathbb {C} \times \mathbb {R} \approx \mathbb {R} ^{3}} .

Because the domains of the branches were glued only along open sets where their values agreed, the branches glue to give a single well-defined function log R : R → C {\displaystyle \log _{R}\colon R\to \mathbb {C} } .^[10] It maps each point ( z , θ ) {\displaystyle (z,\theta )} on R {\displaystyle R} to ln ⁡ | z | + i θ {\displaystyle \ln |z|+i\theta } . This process of extending the original branch Log {\displaystyle {\text{Log}}} by gluing compatible holomorphic functions is known as analytic continuation.

There is a "projection map" from R {\displaystyle R} down to C ∗ {\displaystyle \mathbb {C} ^{*}} that "flattens" the spiral, sending ( z , θ ) {\displaystyle (z,\theta )} to z {\displaystyle z} . For any z ∈ C ∗ {\displaystyle z\in \mathbb {C} ^{*}} , if one takes all the points ( z , θ ) {\displaystyle (z,\theta )} of R {\displaystyle R} lying "directly above" z {\displaystyle z} and evaluates log R {\displaystyle \log _{R}} at all these points, one gets all the logarithms of z {\displaystyle z} .

Instead of gluing only the branches chosen above, one can start with all branches of log ⁡ z {\displaystyle \log z} , and simultaneously glue every pair of branches L 1 : U 1 → C {\displaystyle L_{1}\colon U_{1}\to \mathbb {C} } and L 2 : U 2 → C {\displaystyle L_{2}\colon U_{2}\to \mathbb {C} } along the largest open subset of U 1 ∩ U 2 {\displaystyle U_{1}\cap U_{2}} on which L 1 {\displaystyle L_{1}} and L 2 {\displaystyle L_{2}} agree. This yields the same Riemann surface R {\displaystyle R} and function log R {\displaystyle \log _{R}} as before. This approach, although slightly harder to visualize, is more natural in that it does not require selecting any particular branches.

If U ′ {\displaystyle U'} is an open subset of R {\displaystyle R} projecting bijectively to its image U {\displaystyle U} in C ∗ {\displaystyle \mathbb {C} ^{*}} , then the restriction of log R {\displaystyle \log _{R}} to U ′ {\displaystyle U'} corresponds to a branch of log ⁡ z {\displaystyle \log z} defined on U {\displaystyle U} . Every branch of log ⁡ z {\displaystyle \log z} arises in this way.

The projection map R → C ∗ {\displaystyle R\to \mathbb {C} ^{*}} realizes R {\displaystyle R} as a covering space of C ∗ {\displaystyle \mathbb {C} ^{*}} . In fact, it is a Galois covering with deck transformation group isomorphic to Z {\displaystyle \mathbb {Z} } , generated by the homeomorphism sending ( z , θ ) {\displaystyle (z,\theta )} to ( z , θ + 2 π ) {\displaystyle (z,\theta +2\pi )} .

As a complex manifold, R {\displaystyle R} is biholomorphic with C {\displaystyle \mathbb {C} } via log R {\displaystyle \log _{R}} . (The inverse map sends z {\displaystyle z} to ( e z , Im ⁡ ( z ) ) {\displaystyle \left(e^{z},\operatorname {Im} (z)\right)} .) This shows that R {\displaystyle R} is simply connected, so R {\displaystyle R} is the universal cover of C ∗ {\displaystyle \mathbb {C} ^{*}} .

• The complex logarithm is needed to define exponentiation in which the base is a complex number. Namely, if a {\displaystyle a} and b {\displaystyle b} are complex numbers with a ≠ 0 {\displaystyle a\not =0} , one can use the principal value to define a b = e b Log ⁡ a {\displaystyle a^{b}=e^{b\operatorname {Log} a}} . One can also replace Log ⁡ a {\displaystyle \operatorname {Log} a} by other logarithms of a {\displaystyle a} to obtain other values of a b {\displaystyle a^{b}} , differing by factors of the form e 2 π i n b {\displaystyle e^{2\pi inb}} .^[1][11] The expression a b {\displaystyle a^{b}} has a single value if and only if b {\displaystyle b} is an integer.^[1]
• Because trigonometric functions can be expressed as rational functions of e i z {\displaystyle e^{iz}} , the inverse trigonometric functions can be expressed in terms of complex logarithms.
• In electrical engineering, the propagation constant involves a complex logarithm.

Just as for real numbers, one can define for complex numbers b {\displaystyle b} and x {\displaystyle x}

log b ⁡ x = log ⁡ x log ⁡ b , {\displaystyle \log _{b}x={\frac {\log x}{\log b}},}

with the only caveat that its value depends on the choice of a branch of log defined at b {\displaystyle b} and x {\displaystyle x} (with log ⁡ b ≠ 0 {\displaystyle \log b\not =0} ). For example, using the principal value gives

log i ⁡ e = Log ⁡ e Log ⁡ i = 1 π i / 2 = − 2 i π . {\displaystyle \log _{i}e={\frac {\operatorname {Log} e}{\operatorname {Log} i}}={\frac {1}{\pi i/2}}=-{\frac {2i}{\pi }}.}

If f is a holomorphic function on a connected open subset U {\displaystyle U} of C {\displaystyle \mathbb {C} } , then a branch of log ⁡ f {\displaystyle \log f} on U {\displaystyle U} is a continuous function g {\displaystyle g} on U {\displaystyle U} such that e g ( z ) = f ( z ) {\displaystyle e^{g(z)}=f(z)} for all z {\displaystyle z} in U {\displaystyle U} . Such a function g {\displaystyle g} is necessarily holomorphic with g ′ ( z ) = f ′ ( z ) / f ( z ) {\displaystyle g'(z)=f'(z)/f(z)} for all z {\displaystyle z} in U {\displaystyle U} .

If U {\displaystyle U} is a simply connected open subset of C {\displaystyle \mathbb {C} } , and f {\displaystyle f} is a nowhere-vanishing holomorphic function on U {\displaystyle U} , then a branch of log ⁡ f {\displaystyle \log f} defined on U {\displaystyle U} can be constructed by choosing a starting point a in U {\displaystyle U} , choosing a logarithm b {\displaystyle b} of f ( a ) {\displaystyle f(a)} , and defining

g ( z ) = b + ∫ a z f ′ ( w ) f ( w ) d w {\displaystyle g(z)=b+\int _{a}^{z}{\frac {f'(w)}{f(w)}}\,dw}

for each z {\displaystyle z} in U {\displaystyle U} .^[2]

• Calkin, Neil J.; Chan, Eunice Y. S.; Corless, Robert M. (2023). Computational Discovery on Jupyter. Society for Industrial and Applied Mathematics. ISBN 978-1-61197-749-3.
• Jeffrey, D. J.; Hare, D. E. G.; Corless, Robert M. (1996). Unwinding the branches of the Lambert W function. The Mathematical Scientist, 21, 1–7.
• Ahlfors, Lars V. (1966). Complex Analysis (2nd ed.). McGraw-Hill.
• Conway, John B. (1978). Functions of One Complex Variable (2nd ed.). Springer. ISBN 9780387903286.
• Kreyszig, Erwin (2011). Advanced Engineering Mathematics (10th ed.). Berlin: Wiley. ISBN 9780470458365.
• Lang, Serge (1993). Complex Analysis (3rd ed.). Springer-Verlag. ISBN 9783642592737.
• Moretti, Gino (1964). Functions of a Complex Variable. Prentice-Hall.
• Sarason, Donald (2007). Complex Function Theory (2nd ed.). American Mathematical Society. ISBN 9780821886229.
• Whittaker, E. T.; Watson, G. N. (1927). A Course of Modern Analysis (Fourth ed.). Cambridge University Press.