2.4: General Exponential and Logarithmic Functions
( \newcommand{\kernel}{\mathrm{null}\,}\)
For a general exponential function y=ax, with a>0, use logarithmic differentiation to find its derivative:
lny = ln(ax) = xlna\ddx(lny) = \ddx(xlna) = lnay′y = lna⇒y′ = y⋅lna
Thus, the derivative of y=ax is: In general, for an exponent of the form u=u(x):
Find the derivative of y=2cosx.
Solution: This is the case where a=2, so:
\dydx=(ln2)2cosx⋅\ddx(cosx) = −(ln2)(sinx)2cosx
Note that any exponential function y=ax can be expressed in terms of the exponential function ex. Since
ax > 0⇒eln(ax) = ax ,
and since ln(ax)=xlna, then: Computers and calculators often use the above formula to calculate ax.
The function y=ax has an inverse for any a>0, except for a=1 (in that case y=1x=1 is just a constant function). To see this, notice that since ax>0 for all x, and lna<0 for 0<a<1, while lna>0 for a>1, then \dydx=(lna)ax is always negative if 0<a<1 and always positive if a>1. Thus, y=ax is a strictly decreasing function if 0<a<1, and it is a strictly increasing function if a>1. The graphs in each case are shown in Figure [fig:expa].
Hence, for any a>0 with a≠1 the function f(x)=ax is one-to-one, so it has an inverse function, called the base \bma logarithm and denoted by f−1(x)=logax. It is often spoken as “log base a of x”. The graphs for a<1 and a>1 are shown in Figure [fig:loga]. Note that the natural logarithm is just the base a logarithm in the special case with a=e, i.e. lnx=logex. The base a logarithm has properties similar to those of the natural logarithm (and the corresponding properties of ax):
Note that logax can be put in terms of the natural logarithm, since
x = alogax⇒lnx = ln(alogax) = (logax)⋅(lna)
so dividing the last expression by lna gives: The above formula is useful on calculators that do not have a logax key or function. Taking the derivative of both sides yields: In general, when taking the logarithm of a function u=u(x):
Find the derivative of y=log2(cos4x).
Solution: This is the case where a=2, so:
\dydx=1(cos4x)(ln2)⋅\ddx(cos4x) = −4sin4x(ln2)(cos4x)
The number a is the base of both the logarithm function logax and the exponential function ax. Base 2 and base 10 are the most commonly used bases other than base e. Base 10 is how numbers are normally expressed, as combinations of powers of 10 (e.g. 2014=\bm2⋅103+\bm0⋅102+\bm1⋅101+\bm4⋅100). Base 2 is especially useful in computer science, since computers represent all numbers in binary format, i.e. as a sequence of zeros and ones, indicating how many successive powers of two to take and then sum up.8 For example, the number 6 is represented in binary format as 110, since \bm1⋅22+\bm1⋅21+\bm0⋅20 = 4+2+0=6.
[sec2dot4]
For Exercises 1-9, find the derivative of the given function.
3
y=3x + 3−x2
y=2ln3x3x2
y=22x3x2
3
y=tan−1πx
y=log2(x2+1)
y=log10ex
3
y=sin(log2πx)
y=log242x
y=8log2x
[[1.]]
Show that for all constants k the function y=Aakxlna satisfies the differential equation \dydx=ky. Does this contradict the statement made in Section 2.3 that the only solution to that differential equation is of the form y=Aekx? Explain your answer.