This material has been adapted from material by Fergus Cooper from the "Essential Mathematics" module of the SABS R³ Center for Doctoral Training.
This course material was developed as part of UNIVERSE-HPC, which is funded through the SPF ExCALIBUR programme under grant number EP/W035731/1
Differentiation 3
YouTube lecture recording from October 2020
The following YouTube video was recorded for the 2020 iteration of the course.
The material is still very similar:
Exponentials and Partial Differentiation
Examples of applying chain rule to the exponential function
y=e−ax
Let u=−ax⇒dxdu=−a.
Thus y=eu and
dudy=eu⇒dxdy=dudy×dxdu=eu×(−1.=−ae−ax.
y=ex2
Then, letting u=x2
dxdex2=dxdy=dudy×dxdu=eu⋅2x=ex2⋅2x.
So an important generalization is:
dxdef(x)=ef(x)f′(x) for any function f(x)
Example with the natural logarithm
y=ln(a−x)2=2ln(a−x)=2lnu.
Let u=(a−x):
⇒dxdu=−1anddudy=u2Thusdxdy=u2×(−1)=a−x−2
This also generalises:
dxdln(f(x))=f(x)f′(x)
The Derivative of ax
By the properties of logarithms and indices we have
ax=(elna)x=e(x⋅lna).
Thus, as we saw above we have:
dxdax=dxde(x⋅lna)=e(x⋅lna)dxd(x⋅lna)=ax⋅lna
Similarly, in general:
dxdaf(x)=af(x)⋅lna⋅f′(x)
Sympy Example
Let's try and use Sympy to demonstrate this:
import sympy as sp
x, a = sp.symbols('x a')# declare the variables x and af = sp.Function('f')# declare a function dependent on another variablesp.diff(a**f(x),x)# write the expression we wish to evaluate
af(x)log(a)dxdf(x)
The Derivative of logax
Recall the conversion formula logax=lnalnx
and note that lna is a constant.
Thus:
x, a = sp.symbols('x a')# declare the variables x and af = sp.Function('f')# declare a function dependent on another variablesp.diff(sp.log(f(x),a),x)# write the expression we wish to evaluate
f(x)log(a)dxdf(x)
Further examples
Product Rule: Let y=x2ex. Then:
dxdy=dxdx2ex=dxdx2⋅ex+x2⋅dxdex=(2x+x2)ex
Quotient Rule: Let y=xex. Then:
dxdy=x2dxdex⋅x−ex⋅dxdx=x2ex⋅x−ex⋅1=x2x−1ex
Chain Rule: y=ex2. Then, letting f(x)=x2:
dxdex2=ef(x)f′(x)=ex2⋅2x
y=ln(x2+1). Then, letting f(x)=x2+1:
dxdln(x2+1)=f(x)f′(x)=x2+12x
dxd2x3=2x3⋅ln2⋅3x2
dxd10x2+1=10x2+1⋅ln10⋅2x
dxdlog10(7x+5)=(7x+5)⋅ln107
dxdlog2(3x+x4)=ln2⋅(3x+x4)3x⋅ln3+4x3
Functions of several variables: Partial Differentiation
Definition: Given a function z=f(x,y) of two variables x and y, the partial derivative of z with respect to x is the function obtained by differentiating f(x,y) with respect to x, holding y constant.
We denote this using ∂ (the "curly" delta, sometimes pronounced "del") as shown below:
∂x∂z=∂x∂f(x,y)=fx(x,y)
Example 1
f(x,y)=z=x2−2y2 > fx=∂x∂z=2xandfy=∂y∂z=−4y
Example 2
Let z=3x2y+5xy2. Then the partial derivative of z with respect to x, holding y fixed, is:
∂x∂z=∂x∂(3x2y+5xy2) > =3y⋅2x+5y2⋅1 > =6xy+5y2
while the partial of z with respect to y holding x fixed is:
∂y∂z=∂y∂(3x2y+5xy2) > =3x2⋅1+5x⋅2y=3x2+10xy
Sympy example
In the previous slide we had:
∂x∂(3x2y+5xy2)=6xy+5y2
Let's redo this in Sympy:
x, y = sp.symbols('x y')sp.diff(3*x**2*y +5*x*y**2,x)
6xy+5y2
Higher-Order Partial Derivatives
Given z=f(x,y) there are now four distinct possibilities for the
second-order partial derivatives.
With respect to x twice:
∂x∂(∂x∂z)=∂x2∂2z=zxx
With respect to y twice:
∂y∂(∂y∂z)=∂y2∂2z=zyy
First with respect to x, then with respect to y:
∂y∂(∂x∂z)=∂y∂x∂2z=zxy
First with respect to y, then with respect to x:
∂x∂(∂y∂z)=∂x∂y∂2z=zyx
Example: LaPlace's equation for equilibrium temperature distribution on a copper plate
Let T(x,y) give the temperature at the point (x,y).
According to a result of the French mathematician Pierre LaPlace (1749 - 1827), at every point (x,y) the second-order partials of T must satisfy the equation:
Txx+Tyy=0
We can verify that the function T(x,y)=y2−x2 satisfies LaPlace's equation:
First with respect to x:
Tx(x,y)=0−2x=−2xsoTxx(x,y)=−2
Then with respect to y:
Ty(x,y)=2y−0=2ysoTyy(x,y)=2
Finally:
Txx(x,y)+Tyy(x,y)=2+(−2)=0
which proves the result.
The function z=x2y−xy2 does not satisfy LaPlace's equation (and so cannot be a model for thermal equilibrium).
First note that
zx=2xy−y2 > zxx=2y
and that
zy=x2−2xy > zyy=−2x
Therefore:
zxx+zyy=2y−2x=0
We can also verify this in Sympy like so:
T1 = y**2- x**2sp.diff(T1, x, x)+ sp.diff(T1, y, y)
0
and for the second function:
T2 = x**2*y - x*y**2sp.diff(T2, x, x)+ sp.diff(T2, y, y)
−2x+2y
A Note on the Mixed Partials fxy and fyx
If all of the partials of f(x,y) exist, then fxy=fyx for all (x,y).
Example
Let z=x2y3+3x2−2y4. Then zx=2xy3+6x and zy=3x2y2−8y3.
Taking the partial of zx with respect to y we get
zxy=∂y∂(2xy3+6x)=6xy2
Taking the partial of zy with respect to x we get the same thing:
zyx=∂x∂(3x2y2−8y3)=6xy2
So the operators ∂x∂ and ∂y∂ are commutative:
i.e.∂x∂(∂y∂z)=∂y∂(∂x∂z)
Introductory problems
Introductory problems 1
Differentiate the following functions with respect to x, using the stated rules where indicated:
Product rule: xex
Product rule: 3x2log(x)
Chain rule: e−4x3
Chain rule: ln(6x3/2)
Chain rule: 10x2
Any rules: 5x−7lnx
Any rules: 2x3−1ex
Any rules: log2(xcos(x))
Introductory problems 2
If y=e−axshowthat2dx2d2y+adxdy−a2y=0.
Introductory problems 3
If y=e−xsin(x)showthatdx2d2y+2dxdy+2y=0.
Main problems
Main problems 1
The power, W, that a certain machine develops is given by the formula
W=EI−RI2
where I is the current and E and R are positive constants.
Find the maximum value of W as I varies.
Main problems 2
Environmental health officers monitoring an outbreak of food poisoning after a wedding banquet were able to model the time course of the recovery of the guests using the equation:
r=1+t100t
where t represents the number of days since infection and r is the percentage of guests who no longer display symptoms.
Determine an expression for the rate of recovery.
Main problems 3
An experiment called 'the reptilian drag race' looks at how agamid lizards accelerate from a standing start.
The distance x travelled in time t on a horizontal surface has been modelled as
x=vmax(t+ke−kt−k1),
where vmax is the maximum velocity, and k is a rate constant.
Find expressions for the velocity v and acceleration a as functions of time.
For vmax=3ms−1 and k=10s−1, sketch x, v and a/10 on the same axes for 0≤t≤1s.
Main problems 4
The distance x that a particular organism travels over time from its starting location is modelled by the equation
x(t)=t2ek(1−t),
where k is a positive constant and 0≤t≤1s.
Sketch x over time for k=21 and k=3.
Calculate an expression for the organism's velocity as a function of time.
What is the largest value of k such that the organism never starts moving back towards where it started?
Main problems 5
The function
S=Smax(1−eτ−t)
is used to describe sediment thickness accumulating in an extensional basin through time.
What is the sedimentation rate?
Extension problems
Extension problems 1
Let z=32x3−43x2y+52y3.
Find zx and zy
Find zxx and zyy
Show that zxy=zyx
Extension problems 2
Show that f∗xy=f_yx for the following functions:
f(x,y)=x2−xy+y3
f(x,y)=eyln(2x−y)
f(x,y)=2xye2xy
f(x,y)=xsin(y)
Extension problems 3
The body mass index, B, is used as a parameter to classify people as underweight, normal, overweight and obese.
It is defined as their weight in kg, w, divided by the square of their height in meters, h.
Sketch a graph of B against w for a person who is 1.7m tall.
Find the rate of change of B with weight of this person.
Sketch a graph of B against h for a child whose weight is constant at 35 kg.
Find the rate of change of B with height h of this child.
Show that (∂h∂w∂2B)=(∂w∂h∂2B).
Extension problems 4
A light wave or a sound wave propagated through time and space can be represented in a simplified form by:
y=Asin(2π(λx+ωt))
where A is the amplitude, λ is the wavelength and ω is the frequency of the wave.
x and t are position and time respectively.
An understanding of this function is essential for many problems such as sound, light microscopy, phase microscopy and X-ray diffraction.
Draw a graph of y as a function of x assuming t=0.
Draw a graph of y as a function of t assuming x=0.
At what values of x and t does the function repeat itself?
Find the rate at which y changes at an arbitrary fixed position.