IB Maths AI notes › Cheat sheet

IB Maths AI cheat sheet

BOOKLET given in the examLEARN you must remember it · Each block links to its lesson.

IB Maths AI

IB MATHS AI · HL
pastpaperprep.com/revision-notes

1 Number and algebra

1.1Standard form

a×10k,1≤a<10,k∈Za \times 10^{k}, \quad 1 \le a < 10, \quad k \in \mathbb{Z}LEARN
To write a number in standard form, put the point after the first non-zero digit. kk counts the places it moved, and is negative for numbers between 00 and 11.LEARN
Multiply: multiply the front numbers, add the powers. Divide: divide, subtract the powers. Integer power nn: (a×10m)n=an×10mn(a \times 10^{m})^{n} = a^{n} \times 10^{mn}. Then rewrite the result so that 1≤a<101 \le a < 10, changing the power to balance.LEARN
Add or subtract only after writing both numbers with the same power of 1010. To compare positive numbers in standard form, compare the powers first.LEARN
Never give calculator notation such as 6.08E-5 as an answer: write 6.08×10−56.08 \times 10^{-5}, with aa exact or to 3 s.f. unless told otherwise.LEARN

1.2Arithmetic sequences and series

un=u1+(n−1)du_{n} = u_{1} + (n - 1)dBOOKLET
Sn=n2(2u1+(n−1)d);Sn=n2(u1+un)S_{n} = \frac{n}{2}\left(2u_{1} + (n - 1)d\right); \quad S_{n} = \frac{n}{2}\left(u_{1} + u_{n}\right)BOOKLET
A sequence is arithmetic when d=un+1−und = u_{n+1} - u_{n} is the same for every pair of consecutive terms. From umu_{m} to unu_{n} there are n−mn - m steps.LEARN
∑r=ab\sum_{r=a}^{b} adds the terms for every whole number rr from aa to bb, both included: b−a+1b - a + 1 terms. A general term mr+cmr + c gives an arithmetic sum with d=md = m.LEARN
Simple interest at R%R\% a year on PP adds P×R100P \times \frac{R}{100} every year, so the yearly balances are arithmetic. If unu_{n} is the balance at the end of year nn, then u1=P+P×R100u_{1} = P + P \times \frac{R}{100}.LEARN
Roughly arithmetic data: d≈last value−first valuenumber of values−1d \approx \frac{\text{last value} - \text{first value}}{\text{number of values} - 1}. Predictions far beyond the data are less reliable.LEARN
A number of terms nn is a positive integer. With a limit, round in the direction the limit needs and check the values either side.LEARN

1.3Geometric sequences and series

A sequence is geometric when every quotient un+1un\frac{u_{n+1}}{u_{n}} equals the same number rr.LEARN
un=u1rn−1u_{n} = u_{1}r^{n-1}BOOKLET
Sn=u1(rn−1)r−1=u1(1−rn)1−r,r≠1S_{n} = \frac{u_{1}(r^{n} - 1)}{r - 1} = \frac{u_{1}(1 - r^{n})}{1 - r}, \quad r \neq 1BOOKLET
Growing by p%p\% per period means r=1+p100r = 1 + \frac{p}{100}; shrinking by p%p\% per period means r=1−p100r = 1 - \frac{p}{100}.LEARN
If you know two terms of a geometric sequence, express each as u1u_{1} times a power of rr, then divide to eliminate u1u_{1}. An even power, e.g. r2=16r^{2} = 16, gives two possible ratios, r=±4r = \pm 4, unless the context rules one out.LEARN
The sum ∑k=ab\sum_{k=a}^{b} contains b−a+1b - a + 1 terms, and its first term comes from k=ak = a.LEARN
For the least whole nn that meets a condition, solve or tabulate on your GDC, check the whole numbers on both sides, and give one value.LEARN
For r>0r > 0 and r≠1r \ne 1, un=u1rn−1u_{n} = u_{1}r^{n-1} is an exponential function of nn. In data, nearly equal ratios suggest a geometric model; nearly equal differences suggest a linear one.LEARN

1.4Compound interest and depreciation

FV=PV×(1+r100k)knFV = PV \times \left(1 + \frac{r}{100k}\right)^{kn}BOOKLET
Compound interest multiplies the balance by the same number each period, so balances are geometric. Simple interest adds the same amount, so balances are arithmetic.LEARN
k=1,2,4,12k = 1, 2, 4, 12 for yearly, half-yearly, quarterly and monthly compounding. The rate per period is rk%\frac{r}{k}\% and the number of periods is knkn.LEARN
FV=PV×(1−r100)nFV = PV \times \left(1 - \frac{r}{100}\right)^{n}LEARN
Finance solver with no regular payments: PMT=0PMT = 0, P/Y=C/Y=kP/Y = C/Y = k, N=knN = kn. Money paid out is negative, money received positive; depreciation has a negative I%I\%.LEARN
Real value, using the real rate: interest rate rr minus inflation rate ii (an approximation): FVreal=PV×(1+r−i100k)knFV_{\text{real}} = PV \times \left(1 + \frac{r - i}{100k}\right)^{kn}.LEARN
When interest is added only at the end of each period, round a number of periods up to reach a target.LEARN

1.5Exponents and introductory logarithms

am×an=am+naman=am−n(am)n=amn(ab)n=anbn(m,n∈Z; a,b≠0 if an exponent is zero or negative)a^{m} \times a^{n} = a^{m+n} \qquad \frac{a^{m}}{a^{n}} = a^{m-n} \qquad \left(a^{m}\right)^{n} = a^{mn} \qquad (ab)^{n} = a^{n}b^{n} \qquad (m, n \in \mathbb{Z};\ a, b \neq 0 \text{ if an exponent is zero or negative})LEARN
For a≠0a \neq 0: a0=1a^{0} = 1 and a−n=1ana^{-n} = \frac{1}{a^{n}}. A negative exponent gives a reciprocal, never a negative number, and 3x−3=3x33x^{-3} = \frac{3}{x^{3}} because only xx carries the exponent.LEARN
ax=b  ⟺  x=log⁡ab,where a>0, b>0, a≠1a^{x} = b \iff x = \log_{a} b, \quad \text{where } a > 0,\ b > 0,\ a \neq 1BOOKLET
You can take the log only of a positive number. The log itself can be negative: for base a>1a > 1, a number between 00 and 11 has a negative log.LEARN
ln⁡x=log⁡ex\ln x = \log_{\mathrm{e}} x for x>0x > 0, where e=2.71828…\mathrm{e} = 2.71828\ldots For b>0b > 0, solve 10x=b10^{x} = b with x=log⁡10bx = \log_{10} b and ex=b\mathrm{e}^{x} = b with x=ln⁡bx = \ln b.LEARN
On the GDC: LOG is base 1010, LN is base e\mathrm{e}, and the log-with-base template does any other base. Write the log you evaluate before the decimal answer.LEARN

1.6Rounding, bounds and percentage error

Significant figures count from the first non-zero digit. Keep place-holding zeros (86 390→86 00086\,390 \to 86\,000 to 22 s.f.) and zeros made by rounding (5.698→5.705.698 \to 5.70 to 33 s.f.).LEARN
Give answers exactly or to 33 s.f. unless told otherwise, and money usually to 22 d.p. Keep full calculator values until the final answer.LEARN
A value rounded to the nearest uu has bounds half a step either side: rounded−u2≤x<rounded+u2\text{rounded} - \frac{u}{2} \le x < \text{rounded} + \frac{u}{2}.LEARN
For positive quantities, a product is largest with every factor at its upper bound, and ab\frac{a}{b} is largest with aa at its upper bound and bb at its lower bound.LEARN
ε=∣vA−vEvE∣×100%\varepsilon = \left|\frac{v_{A} - v_{E}}{v_{E}}\right| \times 100\%BOOKLET
Maximum percentage error from a rounded measurement: find the result at each bound, use each as vEv_{E} with the rounded result as vAv_{A}, and keep the larger percentage error.LEARN
Estimate by rounding each number to 11 s.f. Reject answers that are impossible in context, such as a negative length.LEARN

1.7Amortization and annuities

In IB exams, payments are made at the end of each period: set PMT: ENDPMT\text{: END}. The last payment earns no interest.LEARN
NN is the number of payments. P/YP/Y is payments per year and C/YC/Y is how many times interest is added per year; they can differ.LEARN
Signs: savings have PMT<0PMT < 0 and FV>0FV > 0. An annuity has PV<0PV < 0, PMT>0PMT > 0 and FV=0FV = 0 when it runs out. A loan has PV>0PV > 0, PMT<0PMT < 0 and FV=0FV = 0 when it is cleared.LEARN
Interest on a loan == total of all payments −- amount borrowed. Interest on savings == final value −- total paid in.LEARN
The number of full withdrawals an annuity can make is the solver's NN rounded down.LEARN
Annuity formula, not examined: FV=PMT×(1+i)N−1iFV = PMT \times \frac{(1 + i)^{N} - 1}{i} is the value just after NN end-of-period payments into an empty account, with interest added once per period at rate ii.LEARN

1.8Solving equations with technology

To solve a system: name each unknown with its units, write one equation per fact, write the system down, then use the GDC's simultaneous equation solver.LEARN
Line each equation up as ax+by=cax + by = c (or ax+by+cz=dax + by + cz = d with three unknowns) before entering it, with 00 for a missing unknown.LEARN
In IB AI exams, a system of linear equations (up to 33 unknowns) always has exactly one solution, and any correct method earns the marks.LEARN
For a curve y=ax2+bx+cy = ax^{2} + bx + c through three given points, substitute each point to get three linear equations in aa, bb and cc.LEARN
The roots of p(x)=0p(x) = 0 are the zeros of pp: the xx-coordinates where y=p(x)y = p(x) meets the xx-axis. The yy-intercept p(0)p(0) is not a zero unless p(0)=0p(0) = 0.LEARN
Rearrange a polynomial equation to p(x)=0p(x) = 0 before using the root finder. A polynomial equation of degree n≥1n \ge 1 has at most nn real roots.LEARN
In context, keep only the roots that fit the model, and say why you reject the others.LEARN

1.9 ◆Laws of logarithms

log⁡axy=log⁡ax+log⁡ay\log_{a} xy = \log_{a} x + \log_{a} yBOOKLET
log⁡axy=log⁡ax−log⁡ay\log_{a} \frac{x}{y} = \log_{a} x - \log_{a} yBOOKLET
log⁡axm=mlog⁡ax\log_{a} x^{m} = m\log_{a} xBOOKLET
The laws need aa, xx and yy positive with a≠1a \neq 1. In exams the base is 1010 or e\mathrm{e}. There is no law for the log of a sum.LEARN
For k>0k > 0: y=kebxy = k\mathrm{e}^{bx} gives ln⁡y=ln⁡k+bx\ln y = \ln k + bx, a straight line of ln⁡y\ln y against xx; y=kxny = kx^{n} with x>0x > 0 gives ln⁡y=ln⁡k+nln⁡x\ln y = \ln k + n\ln x, a straight line of ln⁡y\ln y against ln⁡x\ln x.LEARN
Multiplying a positive quantity by c>0c > 0 adds log⁡10c\log_{10} c to its base-1010 log. Multiplying a sound's intensity by cc adds 10log⁡10c10\log_{10} c dB.LEARN
After solving a log equation, check each value in the original equation and reject any value that makes the inside of a log zero or negative.LEARN

1.10 ◆Rational exponents

amn=(an)m=amn,a>0, m∈Z, n∈Z+a^{\frac{m}{n}} = \left(\sqrt[n]{a}\right)^{m} = \sqrt[n]{a^{m}}, \quad a > 0,\ m \in \mathbb{Z},\ n \in \mathbb{Z}^{+}LEARN
For a>0a > 0, a−p=1apa^{-p} = \frac{1}{a^{p}}, so x−12=1xx^{-\frac{1}{2}} = \frac{1}{\sqrt{x}} for x>0x > 0. A negative exponent gives a reciprocal, never a negative number.LEARN
ap×aq=ap+qapaq=ap−q(ap)q=apq(ab)p=apbp(a,b>0, p,q∈Q)a^{p} \times a^{q} = a^{p+q} \qquad \frac{a^{p}}{a^{q}} = a^{p-q} \qquad \left(a^{p}\right)^{q} = a^{pq} \qquad (ab)^{p} = a^{p}b^{p} \qquad (a, b > 0,\ p, q \in \mathbb{Q})LEARN
In a model y=kxpy = kx^{p} with kk constant and x>0x > 0, multiplying xx by c>0c > 0 multiplies yy by cpc^{p}.LEARN
For x>0x > 0, c>0c > 0 and non-zero integers mm and nn, xmn=cx^{\frac{m}{n}} = c gives x=cnmx = c^{\frac{n}{m}}. On the GDC, bracket the exponent, as in 29.5^(2/3), and keep it as a fraction.LEARN

1.11 ◆Infinite geometric series

S∞=u11−r,∣r∣<1S_{\infty} = \frac{u_{1}}{1 - r}, \quad \lvert r\rvert < 1BOOKLET
A geometric series has a sum to infinity only when −1<r<1-1 < r < 1. Then S∞=lim⁡n→∞SnS_{\infty} = \lim_{n \to \infty} S_{n}.LEARN
Working back from S∞S_{\infty} can give two ratios. Reject any with ∣r∣≥1\lvert r\rvert \ge 1, and any the context rules out, and say why.LEARN
A starting amount plus increases with ratio 0<r<10 < r < 1 never exceeds the starting amount plus S∞S_{\infty} of the increases.LEARN
Least nn for a total of positive terms to pass a target: if 0<r<10 < r < 1, first check the target is below S∞S_{\infty}, or it is never passed. Then solve Sn>targetS_{n} > \text{target} on the GDC, test the whole numbers either side and give the first that works.LEARN
Fractals: positive amounts added stage by stage with ratio ∣r∣<1\lvert r\rvert < 1 have a finite total; with r≥1r \ge 1, like the Koch perimeter's 43\frac{4}{3}, they grow without limit.LEARN

1.12 ◆Complex numbers in Cartesian form

z=a+biz = a + b\mathrm{i}BOOKLET
i2=−1\mathrm{i}^{2} = -1. For z=a+biz = a + b\mathrm{i} with a,ba, b real: Re(z)=a\mathrm{Re}(z) = a, Im(z)=b\mathrm{Im}(z) = b (a real number) and z∗=a−biz^{*} = a - b\mathrm{i}.LEARN
Divide by hand: multiply the top and the bottom by the conjugate of the bottom, using zz∗=a2+b2zz^{*} = a^{2} + b^{2}. Find powers with the GDC in a+bia + b\mathrm{i} mode.LEARN
∣z∣=a2+b2\lvert z\rvert = \sqrt{a^{2} + b^{2}}LEARN
arg⁡z\arg z (z≠0z \neq 0): the anticlockwise angle from the positive real axis, in radians, usually −π<arg⁡z≤π-\pi < \arg z \le \pi. Sketch the quadrant: arctan⁡ba\arctan\frac{b}{a} alone is right only when a>0a > 0.LEARN
Δ=b2−4ac\Delta = b^{2} - 4acBOOKLET
If aa, bb, cc are real, a≠0a \neq 0 and Δ<0\Delta < 0: the graph of y=ax2+bx+cy = ax^{2} + bx + c never meets the xx-axis, and the roots are a conjugate pair −b2a±4ac−b22ai-\frac{b}{2a} \pm \frac{\sqrt{4ac - b^{2}}}{2a}\mathrm{i}, with real part at the vertex's xx-coordinate.LEARN

1.13 ◆Complex numbers in polar and exponential form

z=r(cos⁡θ+isin⁡θ)=reiθ=r cis θz = r(\cos\theta + \mathrm{i}\sin\theta) = r\mathrm{e}^{\mathrm{i}\theta} = r\,\mathrm{cis}\,\thetaBOOKLET
For z=a+bi≠0z = a + b\mathrm{i} \ne 0: r=a2+b2r = \sqrt{a^{2} + b^{2}}. For θ\theta, sketch the point: arctan⁡ba\arctan\frac{b}{a} alone points the wrong way when a<0a < 0. Use −π<θ≤π-\pi < \theta \le \pi, in radians, unless told otherwise.LEARN
Multiply: multiply the moduli, add the arguments. Divide: divide the moduli, subtract the arguments. For an integer nn, (reiθ)n=rneinθ\left(r\mathrm{e}^{\mathrm{i}\theta}\right)^{n} = r^{n}\mathrm{e}^{\mathrm{i}n\theta}. Then add or subtract multiples of 2π2\pi to bring the argument into range.LEARN
Add and subtract in Cartesian form, like vectors. ∣z−w∣\lvert z - w\rvert is the distance between the points zz and ww.LEARN
Multiplying by w=Reiφw = R\mathrm{e}^{\mathrm{i}\varphi} (R>0R > 0) is an enlargement with scale factor RR, centre O, and a rotation through φ\varphi about O: anticlockwise if φ>0\varphi > 0, clockwise if φ<0\varphi < 0.LEARN
Same ω\omega only: A1cos⁡(ωt+α1)+A2cos⁡(ωt+α2)=Rcos⁡(ωt+B)A_{1}\cos(\omega t + \alpha_{1}) + A_{2}\cos(\omega t + \alpha_{2}) = R\cos(\omega t + B), where ReiB=A1eiα1+A2eiα2R\mathrm{e}^{\mathrm{i}B} = A_{1}\mathrm{e}^{\mathrm{i}\alpha_{1}} + A_{2}\mathrm{e}^{\mathrm{i}\alpha_{2}}. The same holds with sin⁡\sin throughout.LEARN

1.14 ◆Matrices, determinants and inverses

A=(abcd)⇒det⁡A=∣A∣=ad−bcA = \begin{pmatrix} a & b\\ c & d \end{pmatrix} \Rightarrow \det A = |A| = ad - bcBOOKLET
A=(abcd)⇒A−1=1det⁡A(d−b−ca), ad≠bcA = \begin{pmatrix} a & b\\ c & d \end{pmatrix} \Rightarrow A^{-1} = \frac{1}{\det A}\begin{pmatrix} d & -b\\ -c & a \end{pmatrix},\ ad \neq bcBOOKLET
Order m×nm \times n means mm rows and nn columns. Add or subtract matrices of the same order element by element; kAkA multiplies every element by kk.LEARN
ABAB exists only when AA has as many columns as BB has rows: (m×n)(n×p)(m \times n)(n \times p) gives m×pm \times p. Each element is a row of AA times a column of BB.LEARN
When the products are defined, (AB)C=A(BC)(AB)C = A(BC) and A(B+C)=AB+ACA(B + C) = AB + AC, but in general AB≠BAAB \neq BA.LEARN
AI=IA=AAI = IA = A for a square matrix AA and the identity II of the same order; A+0=AA + 0 = A for the zero matrix of the same order.LEARN
A square matrix AA has an inverse only when det⁡A≠0\det A \neq 0; then AA−1=A−1A=IAA^{-1} = A^{-1}A = I.LEARN
nn linear equations in nn unknowns form Ax=bA\boldsymbol{x} = \boldsymbol{b}. If det⁡A≠0\det A \neq 0, the solution is x=A−1b\boldsymbol{x} = A^{-1}\boldsymbol{b}, with A−1A^{-1} on the left.LEARN
Coding: send C=KMC = KM, with the letter numbers in the columns of MM. Decoding: M=K−1CM = K^{-1}C, which needs det⁡K≠0\det K \neq 0.LEARN

1.15 ◆Eigenvalues and eigenvectors

For a square matrix AA, λ\lambda is an eigenvalue with eigenvector v\boldsymbol{v} when Av=λvA\boldsymbol{v} = \lambda\boldsymbol{v} and v≠0\boldsymbol{v} \neq \boldsymbol{0}. Any non-zero multiple of v\boldsymbol{v} is also an eigenvector for λ\lambda.LEARN
det⁡(A−λI)=λ2−(a+d)λ+(ad−bc)for A=(abcd); its roots are the eigenvalues\det(A - \lambda I) = \lambda^{2} - (a + d)\lambda + (ad - bc) \quad \text{for } A = \begin{pmatrix}a & b\\ c & d\end{pmatrix}\text{; its roots are the eigenvalues}LEARN
To find an eigenvector for an eigenvalue λ\lambda of a 2×22 \times 2 matrix AA, solve (A−λI)v=0(A - \lambda I)\boldsymbol{v} = \boldsymbol{0}. The two equations are multiples of each other: pick any non-zero solution.LEARN
A 2×22 \times 2 matrix with two different real eigenvalues can be written A=PDP−1A = PDP^{-1}: eigenvectors as the columns of PP, and the matching eigenvalues, in the same order, on the diagonal of DD.LEARN
M=PDP−1⇒Mn=PDnP−1,Dn=(λ1 n00λ2 n)M = PDP^{-1} \Rightarrow M^{n} = PD^{n}P^{-1}, \quad D^{n} = \begin{pmatrix}\lambda_{1}^{\,n} & 0\\ 0 & \lambda_{2}^{\,n}\end{pmatrix}BOOKLET
In xn=Mnx0\boldsymbol{x}_n = M^{n}\boldsymbol{x}_0 with M=PDP−1M = PDP^{-1}, the part of x0\boldsymbol{x}_0 along each eigenvector is multiplied by its eigenvalue at every step. As n→∞n \to \infty that part dies away if ∣λ∣<1\lvert\lambda\rvert < 1 and stays the same if λ=1\lambda = 1. If it is not zero and ∣λ∣>1\lvert\lambda\rvert > 1, it grows without limit.LEARN

Which method when

Same amount added each step (equal differences, simple interest), 'find the 20th term' or 'total after n weeks' → arithmetic sequence: un=u1+(n−1)du_{n} = u_{1} + (n - 1)d and Sn=n2(2u1+(n−1)d)S_{n} = \frac{n}{2}(2u_{1} + (n - 1)d) (1.2)LEARN
Grows or falls by a fixed percentage each period, or equal ratios between terms → geometric sequence with r=1±p100r = 1 \pm \frac{p}{100}: un=u1rn−1u_{n} = u_{1}r^{n-1}, and SnS_{n} (r≠1r \ne 1) for a total; for 'first n that…', solve or tabulate on the GDC and check both neighbours (1.3)LEARN
One lump sum left to grow, compounded k times a year, or depreciation at a fixed r% a year → FV=PV(1+r100k)knFV = PV\left(1 + \frac{r}{100k}\right)^{kn} or the finance solver with PMT=0PMT = 0, P/Y=C/Y=kP/Y = C/Y = k, N=knN = kn, PVPV negative and FVFV positive; depreciation uses I%=−rI\% = -r with k=1k = 1 (1.4)LEARN
'In today's prices' or 'allowing for inflation' → real value: use the real rate r−ir - i (an approximation) in the compound interest formula or the finance solver (1.4)LEARN
A regular payment each period: saving monthly, an annuity paying out, or repaying a loan → finance solver with PMTPMT entered, NN = number of payments, payments at the end of each period, money paid out negative (1.7)LEARN
'Maximum percentage error' caused by rounding a measurement → find the result at the lower and upper bounds, use each as vEv_{E} with the rounded result as vAv_{A}, and keep the larger ∣vA−vEvE∣×100%\left|\frac{v_{A} - v_{E}}{v_{E}}\right| \times 100\% (1.6)LEARN
Unknown in the exponent, such as 10x=b10^{x} = b or ex=b\mathrm{e}^{x} = b with b>0b > 0 → x=log⁡10bx = \log_{10} b or x=ln⁡bx = \ln b, or the GDC's equation solver; get the power on its own first (1.5)LEARN
Several facts about the same 2 or 3 unknowns, or a quadratic through three given points → write the system, then the GDC's simultaneous equation solver; a polynomial equation → rearrange to p(x)=0p(x) = 0, use the root finder and keep only the roots that fit the context (1.8)LEARN

◆Which method when

'Write as a single logarithm', 'expand', or a log equation such as ln⁡x+ln⁡(x+5)=ln⁡14\ln x + \ln(x + 5) = \ln 14 → laws of logarithms (booklet; base 1010 or e\mathrm{e}, every inside positive): product → sum, quotient → difference, power → multiple, no law for log⁡(x+y)\log(x + y); reject any solution that makes a log undefined (1.9)LEARN
Fractional powers: 'exact value of 81−3481^{-\frac{3}{4}}', 'write in the form kxpkx^{p}', or solve xmn=cx^{\frac{m}{n}} = c → for a>0a > 0, amn=(an)ma^{\frac{m}{n}} = \left(\sqrt[n]{a}\right)^{m} (root first), laws of exponents with positive bases; for x,c>0x, c > 0, x=cnmx = c^{\frac{n}{m}}, with the exponent bracketed on the GDC (1.10)LEARN
'Total in the long run', 'the most a total of positive terms can reach', or a fractal's area → S∞=u11−rS_{\infty} = \frac{u_{1}}{1 - r}, only if ∣r∣<1\lvert r\rvert < 1, approached but never reached, plus any starting amount; 'first nn when the total passes…' → check the target is below S∞S_{\infty}, then SnS_{n} on the GDC and test both neighbours (1.11)LEARN
Quadratic with real coefficients and b2−4ac<0b^{2} - 4ac < 0, or add, subtract, multiply or divide complex numbers → Cartesian form: quadratic formula with Δ=4ac−b2 i\sqrt{\Delta} = \sqrt{4ac - b^{2}}\,\mathrm{i} (a conjugate pair); divide by hand by multiplying top and bottom by the conjugate of the bottom; powers on the GDC in a+bia + b\mathrm{i} mode (1.12)LEARN
Products, quotients or powers of complex numbers, or 'describe fully' z↦wzz \mapsto wz → exponential form: multiply moduli and add arguments (scale factor ∣w∣\lvert w\rvert, rotation arg⁡w\arg w about O), then bring the argument back into range; adding A1cos⁡(ωt+α1)+A2cos⁡(ωt+α2)A_{1}\cos(\omega t + \alpha_{1}) + A_{2}\cos(\omega t + \alpha_{2}) with the same ω\omega → A1eiα1+A2eiα2A_{1}\mathrm{e}^{\mathrm{i}\alpha_{1}} + A_{2}\mathrm{e}^{\mathrm{i}\alpha_{2}} on the GDC in radians: modulus = amplitude, argument = phase (1.13)LEARN
Tables of amounts and prices, nn linear equations in nn unknowns 'using an inverse matrix', or coding a message → matrices: ABAB needs columns of AA = rows of BB, and order matters; Ax=bA\boldsymbol{x} = \boldsymbol{b} gives x=A−1b\boldsymbol{x} = A^{-1}\boldsymbol{b} when det⁡A≠0\det A \neq 0; decode with M=K−1CM = K^{-1}C (1.14)LEARN
Two groups or places swapping members each step, 'MnM^{n} in terms of nn' or 'long-term' for a 2×22 \times 2 matrix → eigenvalues from det⁡(M−λI)=0\det(M - \lambda I) = 0, eigenvectors from (M−λI)v=0(M - \lambda I)\boldsymbol{v} = \boldsymbol{0}, then Mn=PDnP−1M^{n} = PD^{n}P^{-1} (two different real eigenvalues); parts with ∣λ∣<1\lvert\lambda\rvert < 1 die away and an eigenvector for λ=1\lambda = 1 gives the steady split (1.15)LEARN

2 Functions

2.1Straight-line equations

m=y2−y1x2−x1m = \frac{y_{2} - y_{1}}{x_{2} - x_{1}}BOOKLET
y=mx+c;ax+by+d=0;y−y1=m(x−x1)y = mx + c; \quad ax + by + d = 0; \quad y - y_{1} = m(x - x_{1})BOOKLET
For ax+by+d=0ax + by + d = 0 with b≠0b \neq 0, the gradient is −ab-\frac{a}{b}. Make yy the subject to see it.LEARN
Horizontal lines are y=cy = c, with m=0m = 0. Vertical lines are x=hx = h: the gradient is undefined, so they have no y=mx+cy = mx + c form.LEARN
yy-intercept: put x=0x = 0. xx-intercept: put y=0y = 0. In a linear model the gradient is a rate (units of yy per unit of xx) and the yy-intercept is the value when x=0x = 0.LEARN
Two non-vertical lines are parallel (or the same line) if m1=m2m_{1} = m_{2}, and perpendicular if m1×m2=−1m_{1} \times m_{2} = -1: flip the gradient and change its sign. A horizontal and a vertical line are perpendicular.LEARN
Real slopes: gradient == vertical rise ÷\div horizontal distance, not the length along the slope. A p%p\% gradient means m=p100m = \frac{p}{100}.LEARN

2.2Function concepts, domain and range

A function gives exactly one output f(x)f(x) for each input xx in its domain.LEARN
Domain: the inputs allowed. A square root needs its inside ≥0\ge 0, a denominator can't be 00, and in a model the context limits the inputs too.LEARN
Range: the outputs over the domain, written with f(x)f(x) or yy. Read it from the graph, checking end points, turning points and any value the graph only approaches, which is not included.LEARN
For a one-to-one ff, f−1f^{-1} undoes ff: if f(a)=bf(a) = b, then f−1(b)=af^{-1}(b) = a. Its graph is the reflection of y=f(x)y = f(x) in y=xy = x. f−1(x)f^{-1}(x) is not 1f(x)\frac{1}{f(x)}.LEARN
Only a one-to-one function has an inverse. The domain of f−1f^{-1} is the range of ff, and the range of f−1f^{-1} is the domain of ff.LEARN
For a one-to-one ff and a value kk in its range, solving f(x)=kf(x) = k gives x=f−1(k)x = f^{-1}(k).LEARN

2.3Graphs of functions

(a,b)(a, b) is on the graph of y=f(x)y = f(x) exactly when aa is in the domain and f(a)=bf(a) = b.LEARN
y-intercept: (0,f(0))x-intercepts: solve f(x)=0y\text{-intercept: } (0, f(0)) \qquad x\text{-intercepts: solve } f(x) = 0LEARN
Sketch: the right shape, key features labelled, not to scale. Draw: accurate, to the grid's scale, joined smoothly.LEARN
Label the axes (with units), intercepts, turning points, end points (filled dot if included, open circle if not) and asymptotes (dashed, with equations).LEARN
GDC: set the window from the domain, find key points to 3 s.f., then copy the shape and labels onto paper.LEARN
(f+g)(x)=f(x)+g(x)(f−g)(x)=f(x)−g(x)(f + g)(x) = f(x) + g(x) \qquad (f - g)(x) = f(x) - g(x)LEARN
The graphs of ff and gg meet where f(x)−g(x)=0f(x) - g(x) = 0. Where f(x)−g(x)>0f(x) - g(x) > 0, ff's graph is above gg's.LEARN

2.4Key features of graphs

y-intercept: (0, f(0))x-intercepts: (a, 0) where f(a)=0y\text{-intercept: } (0,\ f(0)) \qquad x\text{-intercepts: } (a,\ 0) \text{ where } f(a) = 0LEARN
Zeros of ff: the solutions of f(x)=0f(x) = 0, the xx-coordinates of the x-intercepts. In a model, reject any outside the domain.LEARN
For a graph with no breaks on a≤x≤ba \le x \le b, the greatest and least values are among the local maximum and minimum values, f(a)f(a) and f(b)f(b): compare them all.LEARN
A parabola with zeros pp and qq has axis of symmetry x=p+q2x = \frac{p + q}{2}, and its vertex lies on it.LEARN
Vertical asymptote x=ax = a: f(x)f(x) increases or decreases without limit as xx gets close to aa. Horizontal asymptote y=by = b: f(x)f(x) gets closer and closer to bb as x→∞x \to \infty or x→−∞x \to -\infty. Write each as an equation.LEARN
Intersections of y=f(x)y = f(x) and y=g(x)y = g(x): solve f(x)=g(x)f(x) = g(x) with the GDC's intersect tool on every crossing, then find each yy from the unrounded xx.LEARN

2.5Modelling with functions

Linear model f(x)=mx+cf(x) = mx + c: mm is the rate of change and c=f(0)c = f(0) the starting value. A piecewise model uses one rule on each interval.LEARN
f(x)=ax2+bx+c  ⇒  axis of symmetry is x=−b2af(x) = ax^{2} + bx + c \;\Rightarrow\; \text{axis of symmetry is } x = -\frac{b}{2a}BOOKLET
Quadratic ax2+bx+cax^{2} + bx + c, a≠0a \ne 0: yy-intercept (0,c)(0, c); the vertex is a maximum if a<0a < 0, a minimum if a>0a > 0. Reject zeros the context rules out.LEARN
For k≠0k \ne 0, kax+cka^{x} + c and ka−x+cka^{-x} + c (with a>0a > 0, a≠1a \ne 1) and kerx+ck\mathrm{e}^{rx} + c (with r≠0r \ne 0) have the horizontal asymptote y=cy = c, and f(0)=k+cf(0) = k + c.LEARN
y=axny = ax^{n} with a≠0a \ne 0 and nn an integer: n>0n > 0 is direct variation (through the origin); n<0n < 0 is inverse variation, with the yy-axis as a vertical asymptote.LEARN
Cubic ax3+bx2+cx+dax^{3} + bx^{2} + cx + d, a≠0a \ne 0: at most two turning points and three zeros; yy-intercept (0,d)(0, d).LEARN
asin⁡(bx)+da\sin(bx) + d or acos⁡(bx)+da\cos(bx) + d, with b>0b > 0 and bxbx in degrees: amplitude ∣a∣=max−min2|a| = \frac{\text{max} - \text{min}}{2}, period 360∘b\frac{360^{\circ}}{b}, principal axis y=d=max+min2y = d = \frac{\text{max} + \text{min}}{2}.LEARN

2.6Fitting and testing models

Modelling process: choose a model type, fit its parameters, test it against data and context, then use it. If the test fails, change the model or its domain.LEARN
For equal steps in xx: equal differences in yy suggest a linear model; equal ratios suggest an exponential model.LEARN
Parameters: an initial condition (t=0t = 0) gives one equation, and each known point gives another. A quadratic through three points with different xx-values gives three linear equations in aa, bb, cc: solve them on the GDC.LEARN
A reasonable domain starts where the situation starts, stops where the model stops making sense (such as a height reaching 00), and uses whole numbers for counts.LEARN
Test a model: compare it with data (percentage error ε=∣vA−vEvE∣×100%\varepsilon = \left|\frac{v_{A} - v_{E}}{v_{E}}\right| \times 100\%, with vAv_{A} the model value and vEv_{E} the actual value) and check its graph's features, such as asymptotes and maxima, against the context.LEARN
Interpolation predicts inside the data range; extrapolation predicts outside it and is less reliable, because the trend may not continue.LEARN

2.7 ◆Composite and inverse functions

(f∘g)(x)=f(g(x))(f \circ g)(x) = f(g(x))LEARN
In (f∘g)(x)(f \circ g)(x), the function next to xx acts first: do gg, then ff. The output of gg must lie in the domain of ff.LEARN
In context, (f∘g)(x)(f \circ g)(x) means “do the step gg models, then the step ff models”. Swapping the order usually changes the result.LEARN
(f∘f−1)(x)=x for x in the domain of f−1(f−1∘f)(x)=x for x in the domain of f(f \circ f^{-1})(x) = x \text{ for } x \text{ in the domain of } f^{-1} \qquad (f^{-1} \circ f)(x) = x \text{ for } x \text{ in the domain of } fLEARN
To find f−1(x)f^{-1}(x) for a one-to-one ff: write y=f(x)y = f(x), swap xx and yy, then make yy the subject. The domain of f−1f^{-1} is the range of ff.LEARN
A function has an inverse only if it is one-to-one. A quadratic becomes one-to-one when its domain is restricted to one side of its vertex (h,k)(h, k): x≥hx \ge h or x≤hx \le h.LEARN
When rearranging gives ± \pm\sqrt{\ }, the restricted domain of ff (the range of f−1f^{-1}) decides the sign.LEARN

2.8 ◆Transformations of graphs

y=f(x−a)+b  :  translation by (ab), so a units right and b units upy = f(x - a) + b \;:\; \text{translation by } \begin{pmatrix} a \\ b \end{pmatrix} \text{, so } a \text{ units right and } b \text{ units up}LEARN
y=−f(x)  :  reflection in the x-axisy=f(−x)  :  reflection in the y-axisy = -f(x) \;:\; \text{reflection in the } x\text{-axis} \qquad y = f(-x) \;:\; \text{reflection in the } y\text{-axis}LEARN
y=pf(x), p>0  :  vertical stretch, scale factor py=f(qx), q>0  :  horizontal stretch, scale factor 1qy = pf(x),\ p > 0 \;:\; \text{vertical stretch, scale factor } p \qquad y = f(qx),\ q > 0 \;:\; \text{horizontal stretch, scale factor } \tfrac{1}{q}LEARN
Outside the bracket acts on yy as it looks. Inside acts on xx the opposite way: x−ax - a moves right, and qxqx divides xx-coordinates by qq.LEARN
A vertical stretch y=pf(x)y = pf(x) keeps the xx-axis invariant; a horizontal stretch y=f(qx)y = f(qx) keeps the yy-axis invariant.LEARN
For y=pf(x)+by = pf(x) + b with p>0p > 0: stretch first, then translate by (0b)\begin{pmatrix} 0 \\ b \end{pmatrix}. Horizontal and vertical changes can be done in either order relative to each other.LEARN
Image of (x0,y0)(x_{0}, y_{0}) on y=pf(qx−c)+by = pf(qx - c) + b, with q≠0q \ne 0: solve qx−c=x0qx - c = x_{0} for xx; the new yy is py0+bpy_{0} + b.LEARN
Asymptotes, turning points and the ends of the domain move with the graph. In a model, say what each transformation means.LEARN

2.9 ◆Further function models

Exponential decay N0e−ktN_{0}\mathrm{e}^{-kt} with N0>0N_{0} > 0 and k>0k > 0: half-life =ln⁡2k= \frac{\ln 2}{k}, the same whatever the starting amount. After nn half-lives, (12)n\left(\tfrac{1}{2}\right)^{n} of it is left.LEARN
f(x)=a+bln⁡xf(x) = a + b\ln x with x>0x > 0 and b≠0b \neq 0: a=f(1)a = f(1), and multiplying xx by mm adds bln⁡mb\ln m. No value at x=0x = 0 and no horizontal asymptote.LEARN
f(x)=asin⁡(b(x−c))+df(x) = a\sin(b(x - c)) + d with b>0b > 0 and the angle in radians (assumed unless a degree sign is shown): amplitude ∣a∣|a|, period 2πb\frac{2\pi}{b}, principal axis y=dy = d, phase shift cc (to the right if c>0c > 0).LEARN
f(x)=L1+Ce−kx, L,k,C>0f(x) = \frac{L}{1 + C\mathrm{e}^{-kx}},\ L, k, C > 0BOOKLET
Logistic L1+Ce−kx\frac{L}{1 + C\mathrm{e}^{-kx}} with L,C,k>0L, C, k > 0: value at x=0x = 0 is L1+C\frac{L}{1 + C}; carrying capacity LL, the horizontal asymptote y=Ly = L, approached but never reached.LEARN
A piecewise model is continuous at a joining point x=px = p when both rules give the same value at pp: substitute pp into each rule, set them equal and solve for the parameter.LEARN

2.10 ◆Logarithmic scales and linearizing data

On a log⁡10\log_{10} scale, equal steps are equal factors. A point a fraction pp of the way from 10a10^{a} to 10a+110^{a+1} shows 10a+p10^{a+p}.LEARN
With a log vertical axis and time across, a straight line means a constant percentage rate of change (growth if it rises, decay if it falls), and a steeper rising line means faster percentage growth.LEARN
If y=kaxy = ka^{x} with k,a>0k, a > 0, then ln⁡y=(ln⁡a)x+ln⁡k\ln y = (\ln a)x + \ln k: straight on a semi-log graph, with a=egradienta = \mathrm{e}^{\text{gradient}} and k=einterceptk = \mathrm{e}^{\text{intercept}}.LEARN
If y=kxny = kx^{n} with k>0k > 0 and x>0x > 0, then ln⁡y=nln⁡x+ln⁡k\ln y = n\ln x + \ln k: straight on a log-log graph, with n=n = gradient and k=einterceptk = \mathrm{e}^{\text{intercept}}.LEARN
With log⁡10\log_{10} in place of ln⁡\ln, undo the logs with powers of 1010: k=10interceptk = 10^{\text{intercept}}, and a=10gradienta = 10^{\text{gradient}} for an exponential model.LEARN
To choose between exponential and power models for positive data, find rr for ln⁡y\ln y against xx and for ln⁡y\ln y against ln⁡x\ln x. Choose the one with ∣r∣|r| closer to 11, and say so.LEARN
Keep the GDC's full gradient and intercept until you have undone the logs. Round only the final parameters and predictions.LEARN

Which method when

Line parallel or perpendicular to a given line → same gradient, or m1×m2=−1m_{1} \times m_{2} = -1, then y−y1=m(x−x1)y - y_{1} = m(x - x_{1}) through the given point; for vertical or horizontal lines use x=hx = h or y=cy = c (2.1)LEARN
Solve f(x)=kf(x) = k or f(x)=g(x)f(x) = g(x), or when are two quantities equal → write the equation down, graph both sides and use the GDC's intersection tool on every crossing; reject solutions outside the domain (2.2, 2.4)LEARN
Range, or greatest and least values on a domain → graph over that domain and compare the end values with every local maximum and minimum (GDC max/min tool); a value the graph only approaches is not included (2.2, 2.4)LEARN
Zeros, xx-intercepts, break-even, or when something runs out → solve f(x)=0f(x) = 0 with the GDC's zero tool or polynomial solver; reject zeros outside the domain (2.3, 2.4)LEARN
Long-term value, or a level the model approaches but never reaches → the horizontal asymptote, as an equation: y=cy = c for kax+cka^{x} + c with k≠0k \ne 0 and 0<a<10 < a < 1 (2.5)LEARN
yy is proportional, or inversely proportional, to a power of xx → y=axny = ax^{n} with nn from the words (inversely as the square: n=−2n = -2), never from whether the data rise or fall; one known pair gives aa (2.5)LEARN
Wave from a maximum, a minimum and the length of one cycle → ∣a∣=max⁡−min⁡2|a| = \frac{\max - \min}{2}, d=max⁡+min⁡2d = \frac{\max + \min}{2}, b=360∘periodb = \frac{360^{\circ}}{\text{period}} with b>0b > 0, and the GDC in degrees (2.5)LEARN
Which type of model? → for equal steps in xx, equal differences in yy suggest linear and equal ratios suggest exponential; a symmetric rise to a peak and fall suggests quadratic; a regular repeat suggests sinusoidal (2.6)LEARN
Parameters from given points → one equation per unknown parameter; for ax2+bx+cax^{2} + bx + c through three points with different xx-values, solve the three linear equations on the GDC (2.6)LEARN

◆Which method when

Find f−1(x)f^{-1}(x) (only a one-to-one ff has one) → write y=f(x)y = f(x), swap xx and yy, rearrange; the domain of f−1f^{-1} is the range of ff. For a quadratic, first restrict to one side of the vertex, and let that domain choose the sign of the square root (2.7)LEARN
Describe how y=f(x)y = f(x) becomes y=pf(qx−c)+by = pf(qx - c) + b, with p,q>0p, q > 0 → vertical stretch, scale factor pp, then translation bb units up; translation cc units right, then horizontal stretch, scale factor 1q\frac{1}{q}. Image of (x0,y0)(x_{0}, y_{0}): solve qx−c=x0qx - c = x_{0}, and y=py0+by = py_{0} + b (2.8)LEARN
Half-life of N0e−ktN_{0}\mathrm{e}^{-kt} with N0,k>0N_{0}, k > 0 → ln⁡2k\frac{\ln 2}{k}, the same from any starting amount; for any other level, solve N(t)=N(t) = that level on the GDC (2.9)LEARN
Restricted growth that levels off at a carrying capacity → logistic L1+Ce−kx\frac{L}{1 + C\mathrm{e}^{-kx}} with L,C,k>0L, C, k > 0: f(0)=L1+Cf(0) = \frac{L}{1 + C} gives CC, and one more point gives kk on the GDC (2.9)LEARN
Exponential or power model for positive data? → find rr for ln⁡y\ln y against xx and for ln⁡y\ln y against ln⁡x\ln x, and choose the one with ∣r∣|r| closer to 11; then y=kaxy = ka^{x} has a=egradienta = \mathrm{e}^{\text{gradient}}, y=kxny = kx^{n} has n=n = gradient, and both have k=einterceptk = \mathrm{e}^{\text{intercept}} (2.10)LEARN

3 Geometry and trigonometry

3.13D distance, midpoint and measurement

Distance in 3D: d=(x1−x2)2+(y1−y2)2+(z1−z2)2d = \sqrt{(x_{1} - x_{2})^{2} + (y_{1} - y_{2})^{2} + (z_{1} - z_{2})^{2}}. The midpoint is the mean of each coordinate.BOOKLET
Right pyramid: V=13AhV = \frac{1}{3}Ah. Right cone: V=13πr2hV = \frac{1}{3}\pi r^{2}h and curved area πrl\pi r l. Sphere: V=43πr3V = \frac{4}{3}\pi r^{3} and A=4πr2A = 4\pi r^{2}.BOOKLET
Right cone: slant height l=r2+h2l = \sqrt{r^{2} + h^{2}}. Use hh for the volume and ll for the curved area.LEARN
Hemisphere: V=23πr3V = \frac{2}{3}\pi r^{3}, curved area 2πr22\pi r^{2}, and total area 3πr23\pi r^{2} when solid. Combined solids: add the volumes and count only the outside surfaces.LEARN
A vertical height meets every horizontal line through its foot at 90∘90^{\circ}. In a square-based right pyramid, the centre of the base to a corner is half the diagonal.LEARN
The angle between a line and a plane is the angle between the line and its projection on the plane, found from a perpendicular dropped to the plane.LEARN

3.2Triangle trigonometry

sin⁡θ=OH,cos⁡θ=AH,tan⁡θ=OA\sin\theta = \frac{\text{O}}{\text{H}}, \quad \cos\theta = \frac{\text{A}}{\text{H}}, \quad \tan\theta = \frac{\text{O}}{\text{A}}LEARN
In a right-angled triangle, find an acute angle by undoing the ratio with sin⁡−1\sin^{-1}, cos⁡−1\cos^{-1} or tan⁡−1\tan^{-1}, with the GDC in degree mode.LEARN
asin⁡A=bsin⁡B=csin⁡C\frac{a}{\sin A} = \frac{b}{\sin B} = \frac{c}{\sin C}BOOKLET
c2=a2+b2−2abcos⁡C;cos⁡C=a2+b2−c22abc^{2} = a^{2} + b^{2} - 2ab\cos C; \quad \cos C = \frac{a^{2} + b^{2} - c^{2}}{2ab}BOOKLET
A=12absin⁡CA = \frac{1}{2}ab\sin CBOOKLET
Sine rule: you know a side and the angle facing it, plus one more side or angle. Cosine rule: you know two sides and the angle between them, or all three sides. Area 12absin⁡C\frac{1}{2}ab\sin C: two sides and the angle between them.LEARN
In a triangle, a negative cosine means an obtuse angle, and sin⁡−1\sin^{-1} never gives an obtuse angle. An angle found with the sine rule is acute when the side facing it is shorter than the side facing the known angle.LEARN

3.3Applied triangle trigonometry

a2+b2=c2,sin⁡θ=OH,cos⁡θ=AH,tan⁡θ=OAa^{2} + b^{2} = c^{2}, \qquad \sin\theta = \frac{\text{O}}{\text{H}}, \quad \cos\theta = \frac{\text{A}}{\text{H}}, \quad \tan\theta = \frac{\text{O}}{\text{A}}LEARN
asin⁡A=bsin⁡B=csin⁡C\frac{a}{\sin A} = \frac{b}{\sin B} = \frac{c}{\sin C}BOOKLET
c2=a2+b2−2abcos⁡C;cos⁡C=a2+b2−c22abc^{2} = a^{2} + b^{2} - 2ab\cos C; \quad \cos C = \frac{a^{2} + b^{2} - c^{2}}{2ab}BOOKLET
bearing of A from B=bearing of B from A±180∘\text{bearing of } A \text{ from } B = \text{bearing of } B \text{ from } A \pm 180^{\circ}LEARN
Draw the diagram first: horizontal ground, vertical lines at right angles, every given value and the unknown labelled.LEARN
In a triangle with no right angle: sine rule with a side, its opposite angle and one more side or angle; cosine rule with two sides and the angle between them, or three sides.LEARN
Elevation and depression are measured from the horizontal. Over horizontal ground, the depression of B from A equals the elevation of A from B.LEARN
A bearing is clockwise from north, with three figures before any decimal point. The back bearing is the bearing ±180∘\pm 180^{\circ}.LEARN
Set the GDC to degrees. Store intermediate values and round only the final answer, with its unit.LEARN

3.4Arc length and sector area

l=θ360×2πrl = \frac{\theta}{360} \times 2\pi rBOOKLET
A=θ360×πr2A = \frac{\theta}{360} \times \pi r^{2}BOOKLET
In both formulas θ\theta is the angle at the centre in degrees and rr is the radius.LEARN
Perimeter of a sector == arc +2r+ 2r: the arc plus the two radii.LEARN
The major sector made by two radii has angle 360∘−θ360^{\circ} - \theta, where θ\theta is the minor sector's angle.LEARN
Segment between a chord and its minor arc == sector −- triangle =θ360×πr2−12r2sin⁡θ= \frac{\theta}{360} \times \pi r^{2} - \frac{1}{2}r^{2}\sin\theta, with θ\theta the angle at the centre in degrees (GDC in degree mode).LEARN

3.5Perpendicular bisectors

The perpendicular bisector of [AB][AB] passes through the midpoint of [AB][AB] at right angles to it. In a plane, a point is on it exactly when it is the same distance from AA as from BB.LEARN
From two points: find the midpoint and the gradient mm of [AB][AB]. If m≠0m \neq 0, the bisector has gradient −1m-\frac{1}{m}; use it with the midpoint in y−y1=m(x−x1)y - y_{1} = m(x - x_{1}).LEARN
From a segment's equation and midpoint: read the segment's gradient mm from the equation (make yy the subject). If m≠0m \neq 0, use −1m-\frac{1}{m} through the midpoint.LEARN
A horizontal segment has a vertical bisector, x=x = the midpoint's xx-coordinate. A vertical segment has a horizontal bisector, y=y = the midpoint's yy-coordinate.LEARN

3.6Voronoi diagrams

Every point inside a site's cell is closer to that site than to any other site.LEARN
An edge lies on the perpendicular bisector of its two sites.LEARN
A vertex is equally far from the sites of all the cells meeting there.LEARN
Edge: through the midpoint of two sites, with gradient −1m-\frac{1}{m} if the segment's gradient is m≠0m \neq 0 (level sites: vertical edge; one above the other: horizontal edge). Vertex: solve two edges simultaneously.LEARN
Adding a site: draw its bisectors with neighbouring sites and erase old edges inside its new cell.LEARN
Nearest neighbour interpolation: every point in a cell gets its site's value.LEARN
Toxic waste dump problem (IB exams): compare each vertex's distance to one of its own sites; the largest wins.LEARN
If 11 unit is kk km on a map, 11 square unit is k2 km2k^2\text{ km}^{2}.LEARN

3.7 ◆Radians, arc length and sector area

π\pi radians =180∘= 180^{\circ}, so a full turn is 2π2\pi radians. Multiply by π180\frac{\pi}{180} to get radians, and by 180π\frac{180}{\pi} to get degrees.LEARN
l=rθl = r\thetaBOOKLET
A=12r2θA = \frac{1}{2}r^{2}\thetaBOOKLET
In l=rθl = r\theta and A=12r2θA = \frac{1}{2}r^{2}\theta, rr is the radius and θ\theta is the angle at the centre in radians.LEARN
Perimeter of a sector =2r+rθ= 2r + r\theta, with θ\theta in radians. The major sector made by two radii has angle 2π−θ2\pi - \theta, where θ\theta is the minor sector's angle.LEARN
Set your GDC to radian mode before finding sin⁡\sin, cos⁡\cos or tan⁡\tan of an angle in radians.LEARN

3.8 ◆Unit-circle trigonometry and equations

On the unit circle, the point at angle θ\theta anticlockwise from the positive xx-axis is (cos⁡θ,sin⁡θ)(\cos\theta, \sin\theta). On a circle of radius rr centred at the origin it is (rcos⁡θ,rsin⁡θ)(r\cos\theta, r\sin\theta).LEARN
sin⁡x\sin x and cos⁡x\cos x stay between −1-1 and 11 and repeat every 360∘360^{\circ} (2π2\pi radians).LEARN
cos⁡2θ+sin⁡2θ=1\cos^{2}\theta + \sin^{2}\theta = 1BOOKLET
tan⁡θ=sin⁡θcos⁡θ\tan\theta = \frac{\sin\theta}{\cos\theta}BOOKLET
tan⁡θ\tan\theta is the gradient of OPOP; it is undefined when cos⁡θ=0\cos\theta = 0. After a square root, choose the sign of sin⁡θ\sin\theta or cos⁡θ\cos\theta from where PP lies; in a triangle sin⁡θ>0\sin\theta > 0.LEARN
Ambiguous case (two sides and an angle not between them): if the sine rule gives sin⁡C=k\sin C = k with 0<k<10 < k < 1, test both C1=sin⁡−1kC_{1} = \sin^{-1}k and 180∘−C1180^{\circ} - C_{1}; keep each one whose sum with the known angle is under 180∘180^{\circ}. If k=1k = 1, C=90∘C = 90^{\circ}; if k>1k > 1 there is no triangle.LEARN
Solve a trigonometric equation in an interval by graphing both sides in the right angle mode over the whole interval, then finding every intersection.LEARN

3.9 ◆Matrix transformations

(k001): horizontal stretch, scale factor k\begin{pmatrix}k & 0\\0 & 1\end{pmatrix}:\ \text{horizontal stretch, scale factor } kBOOKLET
(100k): vertical stretch, scale factor k\begin{pmatrix}1 & 0\\0 & k\end{pmatrix}:\ \text{vertical stretch, scale factor } kBOOKLET
(k00k): enlargement, scale factor k, centre (0,0)\begin{pmatrix}k & 0\\0 & k\end{pmatrix}:\ \text{enlargement, scale factor } k,\ \text{centre } (0, 0)BOOKLET
(cos⁡θ−sin⁡θsin⁡θcos⁡θ): anticlockwise rotation of angle θ about the origin (θ>0)\begin{pmatrix}\cos\theta & -\sin\theta\\\sin\theta & \cos\theta\end{pmatrix}:\ \text{anticlockwise rotation of angle } \theta \text{ about the origin } (\theta > 0)BOOKLET
(cos⁡θsin⁡θ−sin⁡θcos⁡θ): clockwise rotation of angle θ about the origin (θ>0)\begin{pmatrix}\cos\theta & \sin\theta\\-\sin\theta & \cos\theta\end{pmatrix}:\ \text{clockwise rotation of angle } \theta \text{ about the origin } (\theta > 0)BOOKLET
(cos⁡2θsin⁡2θsin⁡2θ−cos⁡2θ): reflection in the line y=(tan⁡θ)x\begin{pmatrix}\cos 2\theta & \sin 2\theta\\\sin 2\theta & -\cos 2\theta\end{pmatrix}:\ \text{reflection in the line } y = (\tan\theta)xBOOKLET
x′=Ax+b\boldsymbol{x}' = A\boldsymbol{x} + \boldsymbol{b}: the matrix acts first, then the translation b\boldsymbol{b}. The columns of AA are the images of (1,0)(1, 0) and (0,1)(0, 1) under AA alone, before the translation.LEARN
The booklet's rotations are about the origin, with θ>0\theta > 0. Its reflection matrix uses 2θ2\theta, where the mirror line y=(tan⁡θ)xy = (\tan\theta)x passes through the origin.LEARN
Transformation AA followed by transformation BB has the single matrix BABA: the first one goes on the right.LEARN
With no translation, applying AA nn times gives xn=Anx0\boldsymbol{x}_{n} = A^{n}\boldsymbol{x}_{0}.LEARN
Area of image =∣det⁡A∣×= \lvert\det A\rvert \times area of object for x′=Ax+b\boldsymbol{x}' = A\boldsymbol{x} + \boldsymbol{b}; translations don't change area. For AA then BB, det⁡(BA)=det⁡B×det⁡A\det(BA) = \det B \times \det A.LEARN
A fractal stage making mm non-overlapping copies of each piece, each with area factor ∣det⁡A∣\lvert\det A\rvert, multiplies the area by m∣det⁡A∣m\lvert\det A\rvert: a geometric sequence.LEARN

3.10 ◆Vectors in 2D and 3D

∣v∣=v12+v22+v32, where v=(v1v2v3)\lvert\boldsymbol{v}\rvert = \sqrt{v_{1}^{2} + v_{2}^{2} + v_{3}^{2}}, \text{ where } \boldsymbol{v} = \begin{pmatrix}v_{1}\\v_{2}\\v_{3}\end{pmatrix}BOOKLET
v=(v1v2v3)=v1i+v2j+v3k\boldsymbol{v} = \begin{pmatrix}v_{1}\\v_{2}\\v_{3}\end{pmatrix} = v_{1}\boldsymbol{i} + v_{2}\boldsymbol{j} + v_{3}\boldsymbol{k}; a base vector that doesn't appear has component 0.LEARN
The resultant of two or more vectors is their sum: add matching components. Forces whose resultant is 0\boldsymbol{0} balance.LEARN
AB→=b−a\overrightarrow{AB} = \boldsymbol{b} - \boldsymbol{a}: subtract the start from the end. The distance between A and B is ∣b−a∣\lvert\boldsymbol{b} - \boldsymbol{a}\rvert.LEARN
If v\boldsymbol{v} and w\boldsymbol{w} are non-zero, they are parallel exactly when w=kv\boldsymbol{w} = k\boldsymbol{v} for some scalar kk; a negative kk means they point in opposite directions.LEARN
For non-zero v\boldsymbol{v}, the unit vector with the same direction is v∣v∣\frac{\boldsymbol{v}}{\lvert\boldsymbol{v}\rvert}, and the vector of length LL in that direction is Lv∣v∣L\frac{\boldsymbol{v}}{\lvert\boldsymbol{v}\rvert}.LEARN
Velocity = speed ×\times the unit vector in the direction of motion. Speed is the magnitude of velocity, never negative.LEARN

3.11 ◆Vector equations of lines

r=a+λb\boldsymbol{r} = \boldsymbol{a} + \lambda\boldsymbol{b}BOOKLET
In r=a+λb\boldsymbol{r} = \boldsymbol{a} + \lambda\boldsymbol{b}, a\boldsymbol{a} is the position vector of any point on the line and b\boldsymbol{b} is any non-zero vector parallel to it. Through P and Q, use b=PQ→=q−p\boldsymbol{b} = \overrightarrow{\mathrm{PQ}} = \boldsymbol{q} - \boldsymbol{p}.LEARN
For r=OP→+λPQ→\boldsymbol{r} = \overrightarrow{\mathrm{OP}} + \lambda\overrightarrow{\mathrm{PQ}}: λ=0\lambda = 0 is P, λ=1\lambda = 1 is Q, and 0≤λ≤10 \le \lambda \le 1 is the segment PQ. With a different point or direction, λ\lambda means something different.LEARN
A point lies on a line only if one value of λ\lambda gives every one of its coordinates.LEARN
x=x0+λl,y=y0+λm,z=z0+λnx = x_{0} + \lambda l, \quad y = y_{0} + \lambda m, \quad z = z_{0} + \lambda nBOOKLET
From parametric form back to r=a+λb\boldsymbol{r} = \boldsymbol{a} + \lambda\boldsymbol{b}: the constants give the point, the coefficients of λ\lambda give the direction, and a coordinate with no λ\lambda term has direction component 00.LEARN
To find where a line meets a condition such as z=0z = 0: solve that coordinate's equation for λ\lambda, then substitute that λ\lambda into every coordinate. If that coordinate has no λ\lambda term, there is no single meeting point.LEARN

3.12 ◆Kinematics with vectors

r=r0+vt\boldsymbol{r} = \boldsymbol{r}_{0} + \boldsymbol{v}tLEARN
For constant velocity v\boldsymbol{v}: r0\boldsymbol{r}_{0} is the position at t=0t = 0, the speed is ∣v∣\lvert\boldsymbol{v}\rvert, and the distance travelled in time tt is ∣v∣t\lvert\boldsymbol{v}\rvert t.LEARN
Relative position of B from A at time tt: AB→=rB(t)−rA(t)\overrightarrow{\mathrm{AB}} = \boldsymbol{r}_{\mathrm{B}}(t) - \boldsymbol{r}_{\mathrm{A}}(t), B's position minus A's at the same tt. Its magnitude is the distance between them, and the objects meet only if it is 0\boldsymbol{0} at some time.LEARN
Closest approach: minimise d(t)=∣rB(t)−rA(t)∣d(t) = \lvert\boldsymbol{r}_{\mathrm{B}}(t) - \boldsymbol{r}_{\mathrm{A}}(t)\rvert over the allowed times with the GDC. For constant velocities, d(t)2d(t)^{2} is a quadratic (or constant, if the velocities are equal): the minimum is at its vertex when that is an allowed time, otherwise at an end of the allowed times. Give the time and the distance.LEARN
If r=f(t)\boldsymbol{r} = f(t) is a motion that starts at t=0t = 0, the same motion starting at t=at = a is r=f(t−a)\boldsymbol{r} = f(t - a), for t≥at \ge a.LEARN
v=drdt\boldsymbol{v} = \frac{\mathrm{d}\boldsymbol{r}}{\mathrm{d}t} and a=dvdt\boldsymbol{a} = \frac{\mathrm{d}\boldsymbol{v}}{\mathrm{d}t}, component by component. Integrating back needs one constant per component, fixed by a known position or velocity.LEARN
Projectile launched from (x0,y0)(x_{0}, y_{0}) with velocity (uxuy)\begin{pmatrix}u_{x}\\u_{y}\end{pmatrix} at t=0t = 0, ignoring air resistance, with yy up and units of metres and seconds: a=(0−9.8)\boldsymbol{a} = \begin{pmatrix}0\\-9.8\end{pmatrix}, so r=(x0+uxty0+uyt−4.9t2)\boldsymbol{r} = \begin{pmatrix}x_{0} + u_{x}t\\y_{0} + u_{y}t - 4.9t^{2}\end{pmatrix}. For uy>0u_{y} > 0, the highest point is where vy=0v_{y} = 0.LEARN
r=(Rcos⁡(ωt)Rsin⁡(ωt))\boldsymbol{r} = \begin{pmatrix}R\cos(\omega t)\\R\sin(\omega t)\end{pmatrix} (R>0R > 0, ω>0\omega > 0, ωt\omega t in radians) is motion on a circle of radius RR about the origin, with speed RωR\omega and period 2πω\frac{2\pi}{\omega}.LEARN

3.13 ◆Scalar and vector products

cos⁡θ=v1w1+v2w2+v3w3∣v∣∣w∣\cos\theta = \frac{v_{1}w_{1} + v_{2}w_{2} + v_{3}w_{3}}{\lvert\boldsymbol{v}\rvert\lvert\boldsymbol{w}\rvert}BOOKLET
For non-zero vectors, v⋅w=0\boldsymbol{v} \cdot \boldsymbol{w} = 0 exactly when they are perpendicular. A positive scalar product means θ<90∘\theta < 90^{\circ}, and a negative one means θ>90∘\theta > 90^{\circ}.LEARN
The angle between two lines is the angle between their direction vectors, taken as acute or right: if you get θ>90∘\theta > 90^{\circ}, use 180∘−θ180^{\circ} - \theta.LEARN
v×w=(v2w3−v3w2v3w1−v1w3v1w2−v2w1), where v=(v1v2v3),  w=(w1w2w3)\boldsymbol{v}\times\boldsymbol{w} = \begin{pmatrix}v_{2}w_{3} - v_{3}w_{2}\\v_{3}w_{1} - v_{1}w_{3}\\v_{1}w_{2} - v_{2}w_{1}\end{pmatrix}, \text{ where } \boldsymbol{v} = \begin{pmatrix}v_{1}\\v_{2}\\v_{3}\end{pmatrix}, \; \boldsymbol{w} = \begin{pmatrix}w_{1}\\w_{2}\\w_{3}\end{pmatrix}BOOKLET
v×w=∣v∣∣w∣sin⁡θ n\boldsymbol{v} \times \boldsymbol{w} = \lvert\boldsymbol{v}\rvert\lvert\boldsymbol{w}\rvert\sin\theta\,\boldsymbol{n} with θ\theta the angle between v\boldsymbol{v} and w\boldsymbol{w}. It is perpendicular to both, since n\boldsymbol{n} is a unit vector at right angles to each, pointing the way the right-hand screw rule gives; it is 0\boldsymbol{0} when they are parallel. Order matters: w×v=−(v×w)\boldsymbol{w} \times \boldsymbol{v} = -(\boldsymbol{v} \times \boldsymbol{w}).LEARN
A=∣v×w∣, where v and w form two adjacent sides of a parallelogramA = \lvert \boldsymbol{v} \times \boldsymbol{w} \rvert, \text{ where } \boldsymbol{v} \text{ and } \boldsymbol{w} \text{ form two adjacent sides of a parallelogram}BOOKLET
A=12∣v×w∣, where v and w are two sides of a triangle from the same vertexA = \frac{1}{2}\lvert \boldsymbol{v} \times \boldsymbol{w} \rvert, \text{ where } \boldsymbol{v} \text{ and } \boldsymbol{w} \text{ are two sides of a triangle from the same vertex}LEARN
along b: a⋅b∣b∣=∣a∣cos⁡θperpendicular to b: ∣a×b∣∣b∣=∣a∣sin⁡θ\text{along } \boldsymbol{b}: \ \frac{\boldsymbol{a} \cdot \boldsymbol{b}}{\lvert\boldsymbol{b}\rvert} = \lvert\boldsymbol{a}\rvert\cos\theta \qquad \text{perpendicular to } \boldsymbol{b}: \ \frac{\lvert \boldsymbol{a} \times \boldsymbol{b} \rvert}{\lvert\boldsymbol{b}\rvert} = \lvert\boldsymbol{a}\rvert\sin\thetaLEARN

3.14 ◆Graph theory basics

∑deg⁡(V)=2×number of edges\sum \deg(V) = 2 \times \text{number of edges}LEARN
Degree counts edge ends at a vertex: a loop adds 22, and each of several edges between the same two vertices counts.LEARN
Weighted graph: each edge carries a distance, time or cost. The drawing is not to scale, so read the weights.LEARN
Simple: no loops and no multiple edges. Complete KnK_n: simple, with every pair of its nn vertices joined, so it has n(n−1)2\frac{n(n-1)}{2} edges.LEARN
Connected: you can travel along edges between any two vertices. Strongly connected (directed graph): you can travel from any vertex to any other following the arrows.LEARN
In a directed graph, the in-degrees add up to the number of edges, and so do the out-degrees.LEARN
Subgraph: vertices and edges taken from GG, with both ends of every kept edge also kept.LEARN
Tree: connected with no cycles. A tree with nn vertices has n−1n - 1 edges, and a connected graph with nn vertices and n−1n - 1 edges is a tree.LEARN

3.15 ◆Adjacency matrices and walks

Adjacency matrix: entry (i,j)(i, j) is the number of edges directly from ii to jj. Symmetric for an undirected graph; for a directed graph, rows are “from”.LEARN
A walk is a sequence of edges, each starting where the last ended; vertices and edges may repeat. Its length is its number of edges.LEARN
entry (i,j) of Ak=number of walks of length k from i to j\text{entry } (i, j) \text{ of } A^{k} = \text{number of walks of length } k \text{ from } i \text{ to } jLEARN
walks of length 1 to k from i to j=entry (i,j) of A+A2+⋯+Ak\text{walks of length } 1 \text{ to } k \text{ from } i \text{ to } j = \text{entry } (i, j) \text{ of } A + A^{2} + \dots + A^{k}LEARN
The leading diagonal of AkA^{k} counts closed walks of length kk. No zero off the diagonal of A+⋯+AkA + \dots + A^{k} means every vertex reaches every other in at most kk steps.LEARN
Weighted adjacency table: entries are weights of direct edges; blank (not 00) means no direct edge.LEARN
Transition matrix of a strongly connected directed graph, or a connected undirected one: TijT_{ij} is the probability of moving from jj to ii, so columns add to 11. With equally likely edges in a simple graph, Tij=1out-degree of jT_{ij} = \frac{1}{\text{out-degree of } j} when j→ij \to i is an edge, and 00 otherwise.LEARN

3.16 ◆Spanning trees, postman and salesman problems

Trail: no repeated edge. Path: no repeated vertex. Circuit: a closed trail. Cycle: a circuit with no repeated vertex except the start, which is also the end.LEARN
Connected graph: all vertices even gives an Eulerian circuit; exactly two odd vertices gives an Eulerian trail between them, but no circuit; more than two odd gives no Eulerian trail.LEARN
Hamiltonian path or cycle: every vertex exactly once. No degree test; find one by trying, or give a reason none exists (a vertex of degree 11 rules out a cycle).LEARN
An MST of a connected weighted graph with nn vertices has n−1n - 1 edges. Kruskal's: cheapest edges first, rejecting cycles. Prim's: grow from one vertex by the cheapest edge to a new vertex; it runs on a table.LEARN
Chinese postman (closed route, every edge, connected graph): least weight == total edge weight ++ least total of shortest routes over the pairings of the odd vertices.LEARN
Travelling salesman: use the table of least distances. Nearest neighbour gives an upper bound; deleted vertex (MST of the rest plus the two lightest edges at the deleted vertex) gives a lower bound. Keep the largest lower and smallest upper bound.LEARN

Which method when

Length or angle inside a solid, or between a line and the ground → find a right-angled triangle (a vertical height, or the line's projection on the plane), then Pythagoras or SOH CAH TOA; two points given as (x,y,z)(x, y, z) → the 3D distance formula, midpoint = mean of each coordinate (3.1)LEARN
Triangle with no right angle: a side and the angle facing it known → sine rule; two sides and the angle between them, or all three sides → cosine rule; area from two sides and the angle between them → 12absin⁡C\frac{1}{2}ab\sin C (3.2)LEARN
'Bearing', 'angle of elevation' or 'depression' in words → sketch first: north lines at every point, bearings clockwise from north, back bearing = bearing ±180∘\pm 180^{\circ}, elevation and depression from the horizontal; then the triangle rules, GDC in degree mode (3.3)LEARN
Slice of a circle with the angle at the centre in degrees → arc θ360×2πr\frac{\theta}{360} \times 2\pi r, sector θ360×πr2\frac{\theta}{360} \times \pi r^{2}, perimeter = arc +2r+ 2r; minor segment = sector −12r2sin⁡θ- \frac{1}{2}r^{2}\sin\theta in degree mode (3.4)LEARN
'Equally far from two places', a Voronoi edge or 'which site is nearest' → perpendicular bisector: through the midpoint, gradient −1m-\frac{1}{m} if m≠0m \neq 0; a vertex: solve two edges on the GDC; 'as far as possible from the nearest site' → in IB exams, the vertex furthest from its own three sites (3.5, 3.6)LEARN

◆Which method when

Angle in radians (no degree sign) → arc l=rθl = r\theta, sector A=12r2θA = \frac{1}{2}r^{2}\theta with θ\theta in radians; radian mode before any sin⁡\sin, cos⁡\cos or tan⁡\tan (3.7)LEARN
Two sides and an angle not between them, 'find the two possible values' → ambiguous case: sine rule gives sin⁡C=k\sin C = k; test sin⁡−1k\sin^{-1}k and 180∘−sin⁡−1k180^{\circ} - \sin^{-1}k, keeping each whose sum with the known angle is under 180∘180^{\circ}; k>1k > 1 means no triangle (3.8)LEARN
'Solve' a trigonometric equation in an interval → graph both sides on the GDC in the matching angle mode, with a window covering the whole interval, and find every intersection (3.8)LEARN
Stretch, enlargement, rotation or reflection of points → the booklet matrix in x′=Ax+b\boldsymbol{x}' = A\boldsymbol{x} + \boldsymbol{b}; AA followed by BB is the single matrix BABA; 'area of the image' → area of object ×∣det⁡A∣\times \lvert\det A\rvert (3.9)LEARN
Line through a point in a given direction, or through P and Q → r=a+λb\boldsymbol{r} = \boldsymbol{a} + \lambda\boldsymbol{b} with b=q−p\boldsymbol{b} = \boldsymbol{q} - \boldsymbol{p}; 'where does it meet the ground' → parametric form: solve z=0z = 0 for λ\lambda, then substitute into every coordinate (3.11)LEARN
Constant velocity, 'where is it at time tt' → r=r0+vt\boldsymbol{r} = \boldsymbol{r}_{0} + \boldsymbol{v}t, speed ∣v∣\lvert\boldsymbol{v}\rvert; 'do they meet' → one tt for every component; 'closest approach' → minimise ∣rB(t)−rA(t)∣\lvert\boldsymbol{r}_{\mathrm{B}}(t) - \boldsymbol{r}_{\mathrm{A}}(t)\rvert over the allowed times on the GDC; velocity in terms of tt → differentiate or integrate each component, one constant each (3.12)LEARN
Angle between two vectors, or 'show that they are perpendicular' → scalar product: cos⁡θ=v⋅w∣v∣∣w∣\cos\theta = \frac{\boldsymbol{v} \cdot \boldsymbol{w}}{\lvert\boldsymbol{v}\rvert\lvert\boldsymbol{w}\rvert}, with both vectors leaving the same point; non-zero vectors are perpendicular exactly when v⋅w=0\boldsymbol{v} \cdot \boldsymbol{w} = 0; between two lines take the acute (or right) angle (3.13)LEARN
Area of a triangle or parallelogram from 3D points, or a direction perpendicular to two non-parallel vectors → vector product of two sides from the same vertex: parallelogram ∣v×w∣\lvert\boldsymbol{v} \times \boldsymbol{w}\rvert, triangle 12∣v×w∣\frac{1}{2}\lvert\boldsymbol{v} \times \boldsymbol{w}\rvert; part of a\boldsymbol{a} along b\boldsymbol{b} → a⋅b∣b∣\frac{\boldsymbol{a} \cdot \boldsymbol{b}}{\lvert\boldsymbol{b}\rvert} (3.13)LEARN
'Number of walks of length kk from ii to jj' → entry (i,j)(i, j) of AkA^{k} on the GDC, rows 'from'; 'at most kk' → A+A2+⋯+AkA + A^{2} + \dots + A^{k}; random surfer or PageRank → transition matrix with columns 'from', each adding to 11 (3.15)LEARN
'Travel every edge exactly once' in a connected graph → count odd vertices: none gives an Eulerian circuit, two give a trail between them, more give neither; 'every edge at least once and back to the start' → Chinese postman: total weight + least-weight pairing of the odd vertices (3.16)LEARN
'Connect every site as cheaply as possible' → minimum spanning tree by Kruskal's or Prim's (n−1n - 1 edges; Prim's runs on a table); 'visit every place and return' → travelling salesman on a table of least distances: nearest neighbour gives an upper bound, deleted vertex a lower bound (3.16)LEARN

4 Statistics and probability

4.1Sampling, data quality and outliers

Population: the whole group a question is about. Sample: the part you collect data from. Counted data is discrete; measured data is continuous, even when rounded.LEARN
Sampling bias comes from how a sample is chosen; a bigger sample chosen the same way doesn't remove it.LEARN
Leave a missing value out and say how many are missing; never record it as 00. Correct a recording error only if you know the true value; otherwise remove it and say so.LEARN
IQR=Q3−Q1\mathrm{IQR} = Q_{3} - Q_{1} (in the formula booklet).BOOKLET
outlier: x<Q1−1.5×IQRorx>Q3+1.5×IQR\text{outlier: } x < Q_{1} - 1.5 \times \mathrm{IQR} \quad \text{or} \quad x > Q_{3} + 1.5 \times \mathrm{IQR}LEARN
A value exactly on a fence is not an outlier. Keep an outlier that is possible in context; correct or remove only an error.LEARN
Simple random: every sample of that size equally likely. Systematic: every kkth of an ordered list, k=Nnk = \frac{N}{n} (rounded down if not whole), random start in the first kk, stop at nn (if nn doesn't divide NN, the end of the list is never chosen). Stratified: random within groups, in proportion. Quota: fixed numbers per group, collector chooses. Convenience: easiest to reach.LEARN
number from a stratum=size of stratumsize of population×sample size\text{number from a stratum} = \frac{\text{size of stratum}}{\text{size of population}} \times \text{sample size}LEARN

4.2Data presentation

Class intervals are inequalities with no gaps and no overlaps, such as 150≤m<175150 \le m < 175 and 175≤m<200175 \le m < 200. A boundary value goes in the class whose inequality it satisfies.LEARN
In a histogram with equal class widths, each bar's height is its frequency and neighbouring bars touch.LEARN
Plot cumulative frequency at each upper class boundary, starting from 00 at the first lower boundary. It counts values at or below a point, so for “more than” subtract the reading from nn.LEARN
On a cumulative frequency graph of nn values, read across from n4\frac{n}{4} for Q1Q_{1}, n2\frac{n}{2} for the median, 3n4\frac{3n}{4} for Q3Q_{3} and k100×n\frac{k}{100} \times n for the kkth percentile.LEARN
IQR=Q3−Q1\mathrm{IQR} = Q_{3} - Q_{1}BOOKLET
Box plot: box from Q1Q_{1} to Q3Q_{3}, a line at the median, whiskers to the most extreme values that are not outliers, and a cross for each outlier.LEARN
Compare two box plots by their medians and their IQRs or ranges, with numbers and in context. A symmetric box with equal whiskers means the data may be normal; a skewed one is unlikely to be.LEARN

4.3Central tendency and dispersion

xˉ=∑i=1kfixin,where n=∑i=1kfi\bar{x} = \frac{\sum_{i=1}^{k} f_{i}x_{i}}{n}, \quad \text{where } n = \sum_{i=1}^{k} f_{i}BOOKLET
Median: the n+12\frac{n+1}{2}th value in order; for even nn, the mean of the middle two. Mode: the most frequent value.LEARN
Grouped data: estimate the mean with mid-interval values a+b2\frac{a+b}{2}. With equal class widths, the modal class is the class with the highest frequency.LEARN
IQR=Q3−Q1\mathrm{IQR} = Q_{3} - Q_{1}BOOKLET
Quartiles of discrete data: use the GDC. Methods differ, so a by-hand value can differ from the GDC's.LEARN
Standard deviation of a data set: GDC only, using σx\sigma_{x}, not sxs_{x}. Variance =σ2= \sigma^{2}: square the unrounded value.LEARN
Every value x→ax+bx \to ax + b: mean →axˉ+b\to a\bar{x} + b, standard deviation →∣a∣σ\to \lvert a\rvert\sigma, variance →a2σ2\to a^{2}\sigma^{2}. Adding a constant leaves σ\sigma unchanged.LEARN

4.4Correlation and regression (y on x)

Describe correlation by direction (positive, negative or zero) and strength (strong, weak or none), in context.LEARN
−1≤r≤1-1 \le r \le 1LEARN
Pearson's rr measures linear correlation only: rr near 00 means no straight-line link, but there may be a curve.LEARN
If ∣r∣\lvert r\rvert is greater than the critical value for that number of pairs and significance level, the linear correlation is significant at that level.LEARN
A line of best fit by eye and the regression line of yy on xx both pass through the mean point (xˉ,yˉ)(\bar{x}, \bar{y}).LEARN
Correlation does not imply causation: a third variable may drive both.LEARN
In the regression line y=ax+by = ax + b, aa is the predicted change in yy for each increase of 11 in xx, and bb is the predicted yy when x=0x = 0, which only means something if x=0x = 0 makes sense and is near the data.LEARN
Use the yy on xx line only to predict yy, for xx inside the data range and when the correlation is strong; it cannot reliably predict xx from yy.LEARN

4.5Probability foundations and expected value

A trial has outcomes; the sample space UU lists them all; an event is a set of outcomes. For two stages, list pairs or draw a table.LEARN
P(A)=n(A)n(U)when the outcomes in U are equally likely\mathrm{P}(A) = \frac{n(A)}{n(U)} \quad \text{when the outcomes in } U \text{ are equally likely}BOOKLET
Every probability is between 00 and 11. From a two-way table with one person chosen at random from everyone in it, divide by the grand total.LEARN
Relative frequency =times the event happensnumber of trials= \frac{\text{times the event happens}}{\text{number of trials}} is an experimental estimate of a probability. More trials usually give a better estimate.LEARN
P(A)+P(A′)=1\mathrm{P}(A) + \mathrm{P}(A') = 1BOOKLET
For "at least one", use 1−P(none)1 - \mathrm{P}(\text{none}).LEARN
Expected number of occurrences =n×P(A)= n \times \mathrm{P}(A) for nn trials. It is a long-run average, so it need not be a whole number.LEARN

4.6Probability diagrams, rules and independence

Fill a Venn diagram from the middle out: the overlap, then the rest of each circle, then the outside. "Or" includes both: A∪BA \cup B means AA, BB or both.LEARN
P(A∪B)=P(A)+P(B)−P(A∩B)\mathrm{P}(A \cup B) = \mathrm{P}(A) + \mathrm{P}(B) - \mathrm{P}(A \cap B)BOOKLET
Mutually exclusive events can't both happen, so P(A∩B)=0\mathrm{P}(A \cap B) = 0. That is not the same as independent.LEARN
P(A∪B)=P(A)+P(B)for mutually exclusive A,B\mathrm{P}(A \cup B) = \mathrm{P}(A) + \mathrm{P}(B) \quad \text{for mutually exclusive } A, BBOOKLET
P(A∣B)=P(A∩B)P(B),P(B)>0\mathrm{P}(A \mid B) = \frac{\mathrm{P}(A \cap B)}{\mathrm{P}(B)}, \quad \mathrm{P}(B) > 0BOOKLET
P(A∩B)=P(B) P(A∣B)\mathrm{P}(A \cap B) = \mathrm{P}(B)\,\mathrm{P}(A \mid B)LEARN
Tree diagrams: multiply along a path, add across paths. Without replacement, the totals and second-stage probabilities change.LEARN
P(A∩B)=P(A) P(B)for independent A,B\mathrm{P}(A \cap B) = \mathrm{P}(A)\,\mathrm{P}(B) \quad \text{for independent } A, BBOOKLET

4.7Discrete random variables and expected value

A discrete random variable XX has a list of possible values xx. Its probability distribution gives P(X=x)\mathrm{P}(X = x) for each one, as a table or a formula.LEARN
∑P(X=x)=1,0≤P(X=x)≤1\sum \mathrm{P}(X = x) = 1, \quad 0 \le \mathrm{P}(X = x) \le 1LEARN
E(X)=∑x P(X=x)\mathrm{E}(X) = \sum x\,\mathrm{P}(X = x)BOOKLET
E(X)\mathrm{E}(X) is the long-run mean of XX. It need not be a value XX can take. Over nn independent repeats of the trial, expect a total of n×E(X)n \times \mathrm{E}(X).LEARN
GDC: one-variable statistics with the values as data and the probabilities as frequencies gives xˉ=E(X)\bar{x} = \mathrm{E}(X); with the probabilities entered as they are, the total frequency should be 11.LEARN
fair game: E(X)=0,where X is the player’s gain (prize minus cost)\text{fair game: } \mathrm{E}(X) = 0, \quad \text{where } X \text{ is the player's gain (prize minus cost)}LEARN

4.8Binomial distribution

X∼B(n,p)X \sim \mathrm{B}(n, p) counts the successes in a fixed number nn of independent trials, each with two outcomes and the same probability pp of success.LEARN
On a GDC, binomial pdf gives P(X=r)\mathrm{P}(X = r) and binomial cdf gives P(X≤k)\mathrm{P}(X \le k).LEARN
For binomial XX (whole numbers only): P(X≥k)=1−P(X≤k−1)\mathrm{P}(X \ge k) = 1 - \mathrm{P}(X \le k - 1) and P(X>k)=1−P(X≤k)\mathrm{P}(X > k) = 1 - \mathrm{P}(X \le k).LEARN
E(X)=np\mathrm{E}(X) = npBOOKLET
Var(X)=np(1−p)\mathrm{Var}(X) = np(1-p)BOOKLET
For X∼B(n,p)X \sim \mathrm{B}(n, p), npnp is the expected number of successes: a long-run average that need not be a whole number. The standard deviation is np(1−p)\sqrt{np(1-p)}.LEARN

4.9Normal distribution

X∼N(μ,σ2):P(X<μ)=P(X>μ)=0.5,P(X<μ−d)=P(X>μ+d)X \sim \mathrm{N}(\mu, \sigma^{2}): \quad \mathrm{P}(X < \mu) = \mathrm{P}(X > \mu) = 0.5, \qquad \mathrm{P}(X < \mu - d) = \mathrm{P}(X > \mu + d)LEARN
X∼N(μ,σ2):P(μ−σ<X<μ+σ)≈0.68P(μ−2σ<X<μ+2σ)≈0.95P(μ−3σ<X<μ+3σ)≈0.997X \sim \mathrm{N}(\mu, \sigma^{2}): \quad \begin{aligned} \mathrm{P}(\mu - \sigma < X < \mu + \sigma) &\approx 0.68 \\ \mathrm{P}(\mu - 2\sigma < X < \mu + 2\sigma) &\approx 0.95 \\ \mathrm{P}(\mu - 3\sigma < X < \mu + 3\sigma) &\approx 0.997 \end{aligned}LEARN
P(X>k)=p  ⟺  P(X<k)=1−p(X continuous)\mathrm{P}(X > k) = p \iff \mathrm{P}(X < k) = 1 - p \quad (X \text{ continuous})LEARN
X∼N(μ,σ2)X \sim \mathrm{N}(\mu, \sigma^{2}): a symmetric bell-shaped curve about μ\mu, with total area 11. Mean, median and mode all equal μ\mu. The second number is the variance.LEARN
For a continuous random variable, P(X=a)=0\mathrm{P}(X = a) = 0, so << and ≤\le give the same probability.LEARN
Normal cdf takes lower bound, upper bound, μ\mu and σ\sigma. Use −1099-10^{99} as the lower bound for less than, and 109910^{99} as the upper bound for more than.LEARN
Inverse normal, with the tail set to left, returns kk from P(X<k)=p\mathrm{P}(X < k) = p. For a middle region of probability pp centred on μ\mu, each tail holds 1−p2\frac{1 - p}{2}.LEARN
Expected number of nn independent items in a region =n×p= n \times p, using the unrounded probability. Give it to 3 s.f.: it need not be a whole number.LEARN

4.10Spearman's rank correlation

Rank each variable separately, both in the same direction. Tied values each get the mean of the ranks they share.LEARN
rsr_s is Pearson's rr calculated on the ranks. On the GDC: enter the two lists of ranks, run linear regression, read rr.LEARN
−1≤rs≤1-1 \le r_s \le 1. rs=1r_s = 1: the rankings are identical. rs=−1r_s = -1: one is the exact reverse of the other.LEARN
Pearson's rr measures only linear correlation. rsr_s measures any monotonic relationship, straight or curved.LEARN
rsr_s is less sensitive to outliers than rr: an extreme value can only take the top or bottom rank.LEARN
A link that rises then falls can give rr and rsr_s near 00, and neither coefficient shows causation.LEARN

4.11Chi-squared tests and the t-test

fe=row total×column totalgrand totalf_e = \frac{\text{row total} \times \text{column total}}{\text{grand total}}LEARN
H0\mathrm{H}_0 is the no-effect claim: independent, fits the claimed distribution, or μ1=μ2\mu_1 = \mu_2 for population means.LEARN
Reject H0\mathrm{H}_0 when the p-value is less than the significance level, or χcalc2\chi^2_{\text{calc}} is greater than the critical value.LEARN
Expected frequencies are what H0\mathrm{H}_0 predicts. The test needs every expected frequency greater than 55: SL exam questions always meet this; at HL you may need to merge classes first.LEARN
χ² tests use the upper tail only. df =(r−1)(c−1)= (r - 1)(c - 1) for an r×cr \times c contingency table; df =n−1= n - 1 for goodness of fit at SL (nn categories).LEARN
SL t-test: pooled two-sample, unpaired samples, normal populations, equal population variances. ≠\neq is two-tailed; >> or << is one-tailed.LEARN
Every conclusion gives the comparison, the decision and its meaning in context.LEARN

4.12 ◆Data collection, reliability and validity

Good survey questions are unbiased (neutral wording) and precise (one meaning, a stated time period). Structured answer choices don't overlap and cover every possible answer. Ask personal questions only if needed, and anonymously.LEARN
Choose variables linked to the question. For a linear model, compare candidates by the size of rr, not its sign. Use only data from the population and conditions the question is about, and say what you left out.LEARN
In IB χ² questions every expected frequency must be greater than 55. Merge a small class with an adjacent class, observed and expected together, until every expected frequency is greater than 55, then count the categories again.LEARN
Goodness of fit: df=k−1−m\text{df} = k - 1 - m, with kk categories after merging and mm parameters estimated from the data. Binomial pp estimated: m=1m = 1; normal μ\mu and σ\sigma estimated: m=2m = 2; parameters given in the question: m=0m = 0.LEARN
Reliability: the same results when the method is repeated under the same conditions. Validity: the method measures what it claims to. Reliable does not mean valid.LEARN
Reliability checks: test–retest (same test, same people, two times) and parallel forms (two equivalent versions, same people). A strong positive correlation between the two sets of scores suggests reliability.LEARN
Validity checks: content (the test covers every part of what it should) and criterion-related (a strong correlation, in the expected direction, with an established measure of the same thing).LEARN

4.13 ◆Non-linear regression and R²

Choose the model type from the context and the scatter diagram (or the form the question names); the GDC fits the parameters. Give them to 3 s.f.LEARN
Residual =y−y^= y - \hat{y}, observed minus predicted: positive above the curve, negative below.LEARN
SSres=∑i=1n(yi−y^i)2SS_{\text{res}} = \sum_{i=1}^{n} \left(y_{i} - \hat{y}_{i}\right)^{2}LEARN
For models fitted to the same data, a smaller SSresSS_{\text{res}} means a closer fit; the least squares curve of a type has the smallest SSresSS_{\text{res}} of that type (the GDC's exponential and power regressions minimise it for ln⁡y\ln y instead).LEARN
R2R^{2} is the proportion of the variation in yy that the model accounts for. For a linear least squares model, R2=r2R^{2} = r^{2}.LEARN
R2R^{2} alone does not choose a model: check the context, the shape and any extrapolation. On the same data, a cubic regression's R2R^{2} is never lower than a quadratic regression's.LEARN

4.14 ◆Linear combinations and unbiased estimates

E(aX+b)=aE(X)+b\mathrm{E}(aX + b) = a\mathrm{E}(X) + bBOOKLET
Var(aX+b)=a2 Var(X)\mathrm{Var}(aX + b) = a^{2}\,\mathrm{Var}(X)BOOKLET
The standard deviation of aX+baX + b is ∣a∣σ\lvert a\rvert\sigma. Adding a constant moves the mean but not the spread.LEARN
Expected values add for any random variables, independent or not: E(a1X1±⋯±anXn)=a1E(X1)±⋯±anE(Xn)\mathrm{E}(a_{1}X_{1} \pm \dots \pm a_{n}X_{n}) = a_{1}\mathrm{E}(X_{1}) \pm \dots \pm a_{n}\mathrm{E}(X_{n}).BOOKLET
Only for independent X1,…,XnX_{1}, \dots, X_{n}: Var(a1X1±⋯±anXn)=a12Var(X1)+⋯+an2Var(Xn)\mathrm{Var}(a_{1}X_{1} \pm \dots \pm a_{n}X_{n}) = a_{1}^{2}\mathrm{Var}(X_{1}) + \dots + a_{n}^{2}\mathrm{Var}(X_{n}). Every term is added, even for a difference. Add variances, never standard deviations.BOOKLET
The total of nn independent items, each with mean μ\mu and variance σ2\sigma^{2}, has mean nμn\mu and variance nσ2n\sigma^{2}. One item scaled by nn, nXnX, has variance n2σ2n^{2}\sigma^{2}.LEARN
From a random sample, xˉ\bar{x} is an unbiased estimate of μ\mu and sn−12=nn−1sn2s_{n-1}^{2} = \frac{n}{n-1}s_{n}^{2} is an unbiased estimate of σ2\sigma^{2}. On the GDC, sn−1s_{n-1} is Sx (sx on the Casio and Nspire).BOOKLET

4.15 ◆Sample mean and the central limit theorem

If X1,…,XnX_{1}, \dots, X_{n} are independent and each is normal, then a1X1±⋯±anXna_{1}X_{1} \pm \dots \pm a_{n}X_{n} is normal. Find its mean and variance with the booklet rules, then enter σ=variance\sigma = \sqrt{\text{variance}} in normal cdf.LEARN
X∼N(μ,σ2)  ⇒  Xˉ∼N(μ,σ2n)for a random sample of size nX \sim \mathrm{N}(\mu, \sigma^{2}) \;\Rightarrow\; \bar{X} \sim \mathrm{N}\left(\mu, \frac{\sigma^{2}}{n}\right) \quad \text{for a random sample of size } nLEARN
For a random sample of size nn from a population with variance σ2\sigma^{2}, Var(Xˉ)=σ2n\mathrm{Var}(\bar{X}) = \frac{\sigma^{2}}{n} and the standard deviation of Xˉ\bar{X} is σn\frac{\sigma}{\sqrt{n}}: four times the sample size halves the spread.LEARN
Central limit theorem: for a random sample of size nn from any population with mean μ\mu and variance σ2\sigma^{2}, Xˉ\bar{X} is approximately N(μ,σ2n)\mathrm{N}\left(\mu, \frac{\sigma^{2}}{n}\right) when nn is large. How large depends on the population; in IB exams n>30n > 30 is enough.LEARN
A normal population gives an exactly normal Xˉ\bar{X} for every nn. A population that is not normal needs a large sample before Xˉ\bar{X} can be treated as normal.LEARN

4.16 ◆Confidence intervals for the mean

xˉ±z∗σn(normal population, σ known)\bar{x} \pm z^{*}\frac{\sigma}{\sqrt{n}} \qquad (\text{normal population, } \sigma \text{ known})LEARN
xˉ±t∗sn−1n(normal population, σ unknown, ν=n−1)\bar{x} \pm t^{*}\frac{s_{n-1}}{\sqrt{n}} \qquad (\text{normal population, } \sigma \text{ unknown, } \nu = n - 1)LEARN
For the mean of a normal population: σ\sigma known, use the normal distribution; σ\sigma unknown, use the t-distribution with n−1n - 1 degrees of freedom and sn−1s_{n-1}, whatever the sample size.LEARN
Normal multipliers (σ\sigma known): z∗=1.645z^{*} = 1.645 for 90%90\%, 1.9601.960 for 95%95\%, 2.5762.576 for 99%99\%.LEARN
A 95%95\% confidence interval comes from a method that captures μ\mu for about 95%95\% of random samples. One interval either contains μ\mu or not.LEARN
A claimed mean inside the interval is plausible (no evidence against it); one outside the interval is evidence against the claim.LEARN
From the same sample, a higher confidence level gives a wider interval. At the same confidence level and standard deviation, a larger sample gives a narrower one.LEARN

4.17 ◆Poisson distribution

X∼Po(m)X \sim \mathrm{Po}(m) models the number of events in a fixed interval of time or space when the events happen independently at a uniform average rate of mm per interval. XX can be 0,1,2,…0, 1, 2, \ldots with no upper limit.LEARN
E(X)=m\mathrm{E}(X) = mBOOKLET
Var(X)=m\mathrm{Var}(X) = mBOOKLET
Because the rate is uniform, a Poisson mean of mm per unit interval becomes mtmt for an interval tt times as long.LEARN
If X∼Po(m1)X \sim \mathrm{Po}(m_{1}) and Y∼Po(m2)Y \sim \mathrm{Po}(m_{2}) are independent, then X+Y∼Po(m1+m2)X + Y \sim \mathrm{Po}(m_{1} + m_{2}). Multiples such as 2X2X and differences such as X−YX - Y are not Poisson.LEARN
Choosing a model: a continuous measurement with a symmetric bell-shaped spread is normal; successes in a fixed number of independent trials with constant pp are binomial; independent events in an interval at a uniform average rate, with no upper limit, are Poisson.LEARN

4.18 ◆Further hypothesis tests and errors

Xˉ∼N(μ0,σ2n) under H0: μ=μ0(normal population, σ known)\bar{X} \sim \mathrm{N}\left(\mu_0, \frac{\sigma^{2}}{n}\right) \text{ under } \mathrm{H}_0\text{: } \mu = \mu_0 \quad (\text{normal population, } \sigma \text{ known})LEARN
Mean of a normal population: z-test when σ\sigma is known, t-test when σ\sigma is unknown, whatever the sample size. IB questions don't ask for a critical region in a t-test: decide with the p-value.LEARN
Matched pairs: take differences dd in a fixed order, then a one-sample t-test with H0\mathrm{H}_0: μd=0\mu_d = 0, assuming the differences are normally distributed.LEARN
In IB AI, binomial and Poisson tests are one-tailed. The critical region has the largest probability under H0\mathrm{H}_0 that is still less than the significance level.LEARN
Correlation: test H0\mathrm{H}_0: ρ=0\rho = 0 with the GDC, assuming a bivariate normal distribution.LEARN
Type I error: rejecting a true H0\mathrm{H}_0. Its probability is the significance level for a normal test with known σ\sigma, and the critical region's probability under H0\mathrm{H}_0 for a binomial or Poisson test.LEARN
Type II error: not rejecting a false H0\mathrm{H}_0. Its probability is P(not in the critical region)\mathrm{P}(\text{not in the critical region}), using a stated true parameter value.LEARN

4.19 ◆Transition matrices and Markov chains

Tns0=sn,where s0 is the initial stateT^{n}\boldsymbol{s}_0 = \boldsymbol{s}_n, \quad \text{where } \boldsymbol{s}_0 \text{ is the initial state}BOOKLET
Transition matrix: TijT_{ij} is the chance of going to state ii at the next step, given the chain is in state jj now, so each column adds up to 11. Columns are “from”, rows are “to”.LEARN
In a transition diagram the probabilities on the arrows leaving any one state, including its loop, add up to 11.LEARN
(Tn)ij(T^{n})_{ij} is the probability of being in state ii after nn steps, starting in state jj. A probability along one fixed path is a product of single-step probabilities instead.LEARN
A Markov chain is regular when some power of TT has all its entries positive. Then the columns of TnT^{n} all approach the steady state, whatever the start.LEARN
Steady state s\boldsymbol{s}: Ts=sT\boldsymbol{s} = \boldsymbol{s} with entries adding up to 11. For a regular chain it is unique: find it from a column of a high power of TT, or exactly by solving Ts=sT\boldsymbol{s} = \boldsymbol{s} with one equation replaced by the sum =1= 1.LEARN
A steady state is an eigenvector of TT for the eigenvalue 11, divided by the sum of its entries. For a regular chain there is only one.LEARN

Which method when

'Is this value an outlier?' → GDC one-variable stats for Q1Q_{1}, Q3Q_{3}; outlier if x<Q1−1.5×IQRx < Q_{1} - 1.5 \times \mathrm{IQR} or x>Q3+1.5×IQRx > Q_{3} + 1.5 \times \mathrm{IQR} (a value on a fence is not one) (4.1)LEARN
'Estimate the median, quartiles or a percentile' from grouped data → cumulative frequency graph, plotted at upper class boundaries; read across from n4\frac{n}{4}, n2\frac{n}{2}, 3n4\frac{3n}{4}, k100×n\frac{k}{100} \times n (4.2)LEARN
'Find the mean and standard deviation' of data or a frequency table → GDC one-variable stats with the frequency list; quote σx\sigma_{x}, not sxs_{x}; grouped data uses mid-interval values, so the mean is an estimate (4.3)LEARN
'Is there linear correlation? Predict yy' → GDC linear regression for rr and y=ax+by = ax + b; significant if ∣r∣\lvert r\rvert is above the given critical value; predict yy only, for xx inside the data, when the correlation is strong (4.4)LEARN
'Given that', or 'or' / 'and' with an overlap → Venn diagram, two-way table or tree; P(A∣B)=P(A∩B)P(B)\mathrm{P}(A \mid B) = \frac{\mathrm{P}(A \cap B)}{\mathrm{P}(B)}; multiply P(A) P(B)\mathrm{P}(A)\,\mathrm{P}(B) only for independent events (4.6)LEARN
A table of values with probabilities, 'find kk', 'is the game fair?' → ∑P(X=x)=1\sum \mathrm{P}(X = x) = 1 and E(X)=∑x P(X=x)\mathrm{E}(X) = \sum x\,\mathrm{P}(X = x); fair when E(X)=0\mathrm{E}(X) = 0 for the player's gain (4.7)LEARN
'Number of successes in nn trials', fixed nn, same pp, independent trials → X∼B(n,p)X \sim \mathrm{B}(n, p); binomial pdf for P(X=r)\mathrm{P}(X = r), binomial cdf for P(X≤k)\mathrm{P}(X \le k); P(X≥k)=1−P(X≤k−1)\mathrm{P}(X \ge k) = 1 - \mathrm{P}(X \le k - 1) (4.8)LEARN
A measurement modelled as X∼N(μ,σ2)X \sim \mathrm{N}(\mu, \sigma^{2}), 'find the probability' → normal cdf with σ\sigma, not σ2\sigma^{2}; 'find the value the top 10%10\% exceed' → inverse normal with left area 0.90.9 (or area 0.10.1 with the tail set to right) (4.9)LEARN
Ranked data, a curved but always rising or falling link, or an outlier → Spearman's rsr_s: rank both variables the same way, tied values take the mean rank, then GDC linear regression on the ranks (4.10)LEARN
Counts in categories → χ² test on the GDC: two categorical variables in a table, 'is X independent of Y?' → test for independence, df =(r−1)(c−1)= (r - 1)(c - 1); one variable against claimed proportions ('equally likely', stated percentages) → goodness of fit, expected == total × claimed proportion, df =n−1= n - 1 for nn categories; every expected frequency must be greater than 55 (4.11)LEARN
Two independent (unpaired) groups of measurements, 'is the mean of A different from / greater than B?' → pooled two-sample t-test on the GDC, H0\mathrm{H}_0: μ1=μ2\mu_1 = \mu_2, assuming normal populations with equal variances; ≠\neq is two-tailed, >> or << one-tailed (4.11)LEARN

◆Which method when

Goodness of fit to a binomial or normal model whose parameters are estimated from the data → merge adjacent classes (observed and expected) until every expected frequency is greater than 55, then df =k−1−m= k - 1 - m (kk classes after merging, mm parameters estimated) (4.12)LEARN
Curved scatter, or a named model such as y=abxy = ab^{x}, y=axby = ax^{b}, quadratic, cubic or sine → GDC regression of that type, parameters to 3 s.f.; compare models on the same data by SSresSS_{\text{res}} or R2R^{2}, together with the context and any extrapolation, never R2R^{2} alone (4.13)LEARN
Probability for a total or difference of independent normal quantities → mean by the E\mathrm{E} rules, variance ∑ai2Var(Xi)\sum a_{i}^{2}\mathrm{Var}(X_{i}) (added even for a difference), then normal cdf with σ=variance\sigma = \sqrt{\text{variance}}; for the mean of a random sample → Xˉ∼N(μ,σ2n)\bar{X} \sim \mathrm{N}\left(\mu, \frac{\sigma^{2}}{n}\right), exactly for a normal population, approximately by the central limit theorem when n>30n > 30 (4.14, 4.15)LEARN
Events in a fixed interval of time or space, happening independently at a uniform average rate, 'how many in…' → X∼Po(m)X \sim \mathrm{Po}(m) with mm scaled to the interval, independent Poisson counts add; a fixed number nn of independent trials with constant pp → binomial instead (4.17)LEARN
Estimating or testing the mean of one normal population → σ\sigma given: Z interval or z-test; σ\sigma unknown: T interval or t-test with sn−1s_{n-1} and n−1n - 1 degrees of freedom, whatever nn; the same subjects measured twice → one-sample t-test on the differences, H0\mathrm{H}_0: μd=0\mu_{d} = 0 (4.16, 4.18)LEARN
A count tested against a claimed proportion or rate → one-tailed binomial B(n,p)\mathrm{B}(n, p) or Poisson Po(m)\mathrm{Po}(m) test, p-value from the tail H1\mathrm{H}_1 points to; 'is there correlation in the population?' → GDC test of H0\mathrm{H}_0: ρ=0\rho = 0, assuming bivariate normal data (4.18)LEARN
States that change step by step, where the next depends only on the current state ('after 3 days', 'in the long run') → transition matrix with columns 'from', each adding up to 11, and sn=Tns0\boldsymbol{s}_n = T^{n}\boldsymbol{s}_0; long run → for a regular chain, the unique steady state Ts=sT\boldsymbol{s} = \boldsymbol{s} with entries adding up to 11 (an eigenvector for eigenvalue 11, scaled), which sn\boldsymbol{s}_n approaches from any start (4.19)LEARN

5 Calculus

5.1Limits and the derivative as a rate of change

lim⁡x→af(x)=L\lim_{x \to a} f(x) = L: f(x)f(x) gets closer and closer to LL as xx gets closer to aa from both sides. The value at aa itself doesn't matter.LEARN
Estimate a limit from values on both sides of aa: a table, a graph or your GDC's table.LEARN
The gradient of a curve at PP is the gradient of the tangent at PP: the limit of the gradients of chords PQPQ as QQ approaches PP.LEARN
dydx=f′(x)when y=f(x)\frac{\mathrm{d}y}{\mathrm{d}x} = f'(x) \quad \text{when } y = f(x)LEARN
The derivative is the gradient function and a rate of change, such as dVdr\frac{\mathrm{d}V}{\mathrm{d}r} or dsdt\frac{\mathrm{d}s}{\mathrm{d}t}. Its units are units of the top variable per unit of the bottom one.LEARN
Where dydx>0\frac{\mathrm{d}y}{\mathrm{d}x} > 0, yy is increasing as xx increases. Where dydx<0\frac{\mathrm{d}y}{\mathrm{d}x} < 0, yy is decreasing.LEARN

5.2Increasing and decreasing functions

If f′(x)>0f'(x) > 0 for every xx in an interval, ff is increasing on that interval: its graph rises from left to right.LEARN
If f′(x)<0f'(x) < 0 for every xx in an interval, ff is decreasing on that interval: its graph falls from left to right.LEARN
f′(x)=0f'(x) = 0 at a stationary point, where the tangent is horizontal. It may be a local maximum, a local minimum or neither.LEARN
On a graph of y=f′(x)y = f'(x), read only the sign: where it is above the horizontal axis ff is increasing, and where it is below, ff is decreasing.LEARN
To find the intervals: find where f′(x)=0f'(x) = 0 or ff or f′f' is undefined, find the sign of f′(x)f'(x) on each piece, then answer with inequalities inside the domain.LEARN

5.3Differentiation of integer-power polynomials

f(x)=xn⇒f′(x)=nxn−1f(x) = x^{n} \Rightarrow f'(x) = nx^{n-1}BOOKLET
For an integer nn, axnax^{n} differentiates to anxn−1anx^{n-1}: multiply by the power, then lower the power by one.LEARN
A constant term differentiates to 00, and axax differentiates to aa.LEARN
Rewrite axn\frac{a}{x^{n}} as ax−nax^{-n} before differentiating. One less than −n-n is −n−1-n - 1.LEARN
ddx(f(x)+g(x))=f′(x)+g′(x)\frac{\mathrm{d}}{\mathrm{d}x}\left(f(x) + g(x)\right) = f'(x) + g'(x)LEARN
Differentiate a sum term by term. Expand brackets and split a fraction with a single-term denominator first; never multiply the derivatives of the factors.LEARN
The gradient of y=f(x)y = f(x) at x=ax = a is f′(a)f'(a). The height of the curve there is f(a)f(a).LEARN

5.4Tangents and normals

Tangent to y=f(x)y = f(x) at x=ax = a: the point is (a,f(a))(a, f(a)) and the gradient is f′(a)f'(a), so y−f(a)=f′(a)(x−a)y - f(a) = f'(a)(x - a).LEARN
mnormal=−1f′(a),f′(a)≠0m_{\text{normal}} = -\frac{1}{f'(a)}, \quad f'(a) \neq 0LEARN
If f′(a)=0f'(a) = 0, the tangent at (a,f(a))(a, f(a)) is y=f(a)y = f(a) and the normal is x=ax = a.LEARN
To find where the tangent to y=f(x)y = f(x) has gradient mm, solve f′(x)=mf'(x) = m for xx, then use ff for yy.LEARN
With a GDC: write down the value it gives for f′(a)f'(a), keep every digit, write the line, and round only the final answer.LEARN

5.5Antidifferentiation, definite integrals and area

∫xn dx=xn+1n+1+C,n≠−1\int x^{n}\,\mathrm{d}x = \frac{x^{n+1}}{n+1} + C,\quad n \neq -1BOOKLET
Integrate a sum one term at a time: axnax^{n} becomes axn+1n+1\frac{ax^{n+1}}{n+1}, for integers n≠−1n \neq -1. Rewrite axn\frac{a}{x^{n}} as ax−nax^{-n} and expand brackets first, and end with one + C+\,C.LEARN
Boundary condition, on a domain with no breaks: integrate with + C+\,C, substitute the known point into the integrated function (not the derivative), solve for CC, then write the full function.LEARN
If F′(x)=f(x)F'(x) = f(x) and ff is continuous for a≤x≤ba \le x \le b, then ∫abf(x) dx=F(b)−F(a)\int_{a}^{b} f(x)\,\mathrm{d}x = F(b) - F(a). Integrating a rate of change gives the total change, in units of the rate times units of xx.LEARN
If f(x)≥0f(x) \ge 0 for a≤x≤ba \le x \le b, the area between y=f(x)y = f(x), the xx-axis and the lines x=ax = a and x=bx = b is A=∫abf(x) dxA = \int_{a}^{b} f(x)\,\mathrm{d}x. Write this integral down before you evaluate it on the GDC.LEARN

5.6Local maximum and minimum points

At a local maximum or local minimum inside the domain of a smooth curve, f′(x)=0f'(x) = 0: the tangent there is horizontal.LEARN
To find local maximum and minimum points, solve f′(x)=0f'(x) = 0 for xx, check whether each is a maximum, a minimum or neither, then find each yy-coordinate from f(x)f(x), not from f′(x)f'(x).LEARN
If f′(a)=0f'(a) = 0 and f′(x)f'(x) changes from ++ to −- as xx increases through aa, there is a local maximum at x=ax = a. From −- to ++ it is a local minimum. With no change of sign it is neither.LEARN
With a GDC, write down f′(x)=0f'(x) = 0, then solve it with the equation solver, or draw y=f′(x)y = f'(x) with the numerical derivative and find its zeros.LEARN
A local maximum or minimum need not be the greatest or least value on the domain. For a graph with no breaks or sharp corners on a≤x≤ba \le x \le b, compare ff at the local maximum and minimum points and at both end points.LEARN

5.7Optimisation problems

To optimise: write the quantity as a function of one variable, solve f′(x)=0f'(x) = 0, justify the maximum or minimum, compare it with the endpoints of the domain, then answer in context with units.LEARN
With two unknowns, use the constraint (a fixed area, volume or amount of material) to write one in terms of the other before you differentiate.LEARN
At a stationary point, f′(x)f'(x) changing from positive to negative means a local maximum; from negative to positive means a local minimum.LEARN
For a function you can differentiate on all of a≤x≤ba \le x \le b, the greatest or least value is at a stationary point inside the domain or at an endpoint, so compare them all.LEARN
If the variable must be a whole number and the optimum is at a stationary point that is not, compare the whole numbers on each side of it.LEARN
With a GDC, write down the function you graph or the equation you solve, keep full values, and round only the final answer.LEARN

5.8Trapezoidal rule

∫aby dx≈12h((y0+yn)+2(y1+y2+⋯+yn−1)),where h=b−an\int_{a}^{b} y\,\mathrm{d}x \approx \frac{1}{2}h\left((y_{0} + y_{n}) + 2(y_{1} + y_{2} + \dots + y_{n-1})\right),\quad \text{where } h = \frac{b - a}{n}BOOKLET
nn intervals of equal width use n+1n + 1 values, and h=b−anh = \frac{b - a}{n}. The end values y0y_{0} and yny_{n} count once; every other value counts twice.LEARN
For y≥0y \ge 0 the rule estimates the area under the curve. For a rate, it estimates the total change, in the rate's units times the xx-units.LEARN
If the curve bends upwards (concave up) across the whole interval, the trapezoidal rule overestimates ∫aby dx\int_{a}^{b} y\,\mathrm{d}x. If it bends downwards (concave down) across the whole interval, it underestimates. If the bending changes, the shape alone can't tell you.LEARN
Percentage error of a trapezoidal estimate: ∣estimate−exactexact∣×100%\left|\frac{\text{estimate} - \text{exact}}{\text{exact}}\right| \times 100\%, where the exact value is the definite integral from your GDC.LEARN
If every height is rounded and the xx-values are exact, using all the upper bounds gives the upper bound of the trapezoidal estimate, and all the lower bounds its lower bound.LEARN

5.9 ◆Differentiation rules and related rates

y=g(u), where u=f(x)⇒dydx=dydu×dudxy = g(u),\ \text{where } u = f(x) \Rightarrow \frac{\mathrm{d}y}{\mathrm{d}x} = \frac{\mathrm{d}y}{\mathrm{d}u} \times \frac{\mathrm{d}u}{\mathrm{d}x}BOOKLET
y=uv⇒dydx=udvdx+vdudxy = uv \Rightarrow \frac{\mathrm{d}y}{\mathrm{d}x} = u\frac{\mathrm{d}v}{\mathrm{d}x} + v\frac{\mathrm{d}u}{\mathrm{d}x}BOOKLET
y=uv⇒dydx=vdudx−udvdxv2y = \frac{u}{v} \Rightarrow \frac{\mathrm{d}y}{\mathrm{d}x} = \frac{v\frac{\mathrm{d}u}{\mathrm{d}x} - u\frac{\mathrm{d}v}{\mathrm{d}x}}{v^{2}}BOOKLET
dAdt=dAdr×drdt\frac{\mathrm{d}A}{\mathrm{d}t} = \frac{\mathrm{d}A}{\mathrm{d}r} \times \frac{\mathrm{d}r}{\mathrm{d}t}LEARN
For xx in radians: xn→nxn−1x^{n} \to nx^{n-1} for any rational nn; sin⁡x→cos⁡x\sin x \to \cos x; cos⁡x→−sin⁡x\cos x \to -\sin x; tan⁡x→1cos⁡2x\tan x \to \frac{1}{\cos^{2} x}; ex→ex\mathrm{e}^{x} \to \mathrm{e}^{x}; ln⁡x→1x\ln x \to \frac{1}{x} for x>0x > 0.BOOKLET
Chain rule in words: differentiate the outside, keep the inside, multiply by the inside's derivative. So, for a constant kk, ekx→kekx\mathrm{e}^{kx} \to k\mathrm{e}^{kx} and sin⁡(kx)→kcos⁡(kx)\sin(kx) \to k\cos(kx) in radians.LEARN
For a product or a quotient, write down uu, vv, dudx\frac{\mathrm{d}u}{\mathrm{d}x} and dvdx\frac{\mathrm{d}v}{\mathrm{d}x} first. The top of the quotient rule starts with vdudxv\frac{\mathrm{d}u}{\mathrm{d}x}.LEARN
Related rates: link the quantities with an equation true at every moment, differentiate, write the chain rule, then substitute the values at the instant.LEARN
A decreasing quantity has a negative rate of change. Give every rate with its units, such as m2 min−1\text{m}^{2}\text{ min}^{-1}.LEARN
To find a maximum or minimum of a model, find f′(x)f'(x) with these rules and solve f′(x)=0f'(x) = 0. Check the sign of f′(x)f'(x) either side, and compare with the values at the end points of the domain.LEARN

5.10 ◆Second derivative and concavity

The second derivative, f′′(x)f''(x) or d2ydx2\dfrac{\mathrm{d}^{2}y}{\mathrm{d}x^{2}}, is the derivative of f′(x)f'(x): differentiate twice. It measures how fast the gradient is changing.LEARN
Second derivative test, at a stationary point x=ax = a (where f′(a)=0f'(a) = 0): f′′(a)>0f''(a) > 0 gives a local minimum and f′′(a)<0f''(a) < 0 a local maximum. If f′′(a)=0f''(a) = 0, use the sign of f′(x)f'(x) on each side instead.LEARN
The graph of ff is concave-up on an interval where f′′(x)>0f''(x) > 0 throughout, and concave-down on an interval where f′′(x)<0f''(x) < 0 throughout.LEARN
A point of inflexion is where the concavity changes. Find it by solving f′′(x)=0f''(x) = 0 (and checking any xx where f′′f'' is undefined but ff exists) and checking that f′′(x)f''(x) changes sign: y=x4y = x^{4} has f′′(0)=0f''(0) = 0 but no point of inflexion. The gradient there need not be 00.LEARN
At a point of inflexion the rate of change f′(x)f'(x) has a local maximum or minimum: in context, the quantity is changing fastest or slowest there compared with nearby values.LEARN
With a GDC, find f′′(x)f''(x) by hand, write down the function you graph, and find where y=f′′(x)y = f''(x) crosses the axis, or where y=f′(x)y = f'(x) has its turning points.LEARN

5.11 ◆Standard integrals and substitution

∫xn dx=xn+1n+1+C,n≠−1\int x^{n}\,\mathrm{d}x = \frac{x^{n+1}}{n+1} + C, \quad n \neq -1BOOKLET
∫1x dx=ln⁡∣x∣+C\int \frac{1}{x}\,\mathrm{d}x = \ln|x| + CBOOKLET
With xx in radians: ∫sin⁡x dx=−cos⁡x+C\int \sin x\,\mathrm{d}x = -\cos x + C, ∫cos⁡x dx=sin⁡x+C\int \cos x\,\mathrm{d}x = \sin x + C and ∫1cos⁡2x dx=tan⁡x+C\int \frac{1}{\cos^{2}x}\,\mathrm{d}x = \tan x + C. Also ∫ex dx=ex+C\int \mathrm{e}^{x}\,\mathrm{d}x = \mathrm{e}^{x} + C.BOOKLET
∫abf(x) dx=F(b)−F(a)\int_{a}^{b} f(x)\,\mathrm{d}x = F(b) - F(a), where FF is any antiderivative of ff and ff is continuous from aa to bb.LEARN
Linear inside: integrate as if ax+bax + b were xx, then divide by aa (for a≠0a \neq 0).LEARN
Inspection: when the integrand is a constant times g′(x) f(g(x))g'(x)\,f(g(x)), integrate the outer function ff with g(x)g(x) kept inside, differentiate your guess, then fix the constant.LEARN
∫g′(x)g(x) dx=ln⁡∣g(x)∣+C\int \frac{g'(x)}{g(x)}\,\mathrm{d}x = \ln|g(x)| + CLEARN
Substitution: let u=g(x)u = g(x) and replace g′(x) dxg'(x)\,\mathrm{d}x by du\mathrm{d}u. For a definite integral, change the limits to uu-values.LEARN

5.12 ◆Areas and volumes of revolution

A=∫ab∣y∣ dxorA=∫ab∣x∣ dyA = \int_{a}^{b} |y|\,\mathrm{d}x \quad \text{or} \quad A = \int_{a}^{b} |x|\,\mathrm{d}yBOOKLET
With the lower limit smaller than the upper, a definite integral counts a region below the xx-axis, or left of the yy-axis, as negative. If the curve crosses the axis inside the interval, split at each crossing and add the sizes.LEARN
For an area against the yy-axis, or a volume about the yy-axis, write xx (or x2x^{2}) in terms of yy, and use yy-values as the limits.LEARN
V=∫abπy2 dxorV=∫abπx2 dyV = \int_{a}^{b} \pi y^{2}\,\mathrm{d}x \quad \text{or} \quad V = \int_{a}^{b} \pi x^{2}\,\mathrm{d}yBOOKLET
In a volume of revolution, π\pi comes from each disc's area, πr2\pi r^{2}. Square the radius, and never put 2π2\pi in front: 2π2\pi or 360∘360^{\circ} is one full turn.LEARN
∫abπy2 dx\int_{a}^{b} \pi y^{2}\,\mathrm{d}x is the volume when the region between the curve and the xx-axis turns once about the xx-axis. If a region doesn't reach the axis of rotation, the solid has a hole: subtract the volume of the hole.LEARN

5.13 ◆Kinematics

v=dsdtv = \frac{\mathrm{d}s}{\mathrm{d}t}LEARN
a=dvdt=d2sdt2=vdvdsa = \frac{\mathrm{d}v}{\mathrm{d}t} = \frac{\mathrm{d}^{2}s}{\mathrm{d}t^{2}} = v\frac{\mathrm{d}v}{\mathrm{d}s}BOOKLET
Dot notation means derivatives with respect to time: x˙=dxdt\dot{x} = \frac{\mathrm{d}x}{\mathrm{d}t} and x¨=d2xdt2\ddot{x} = \frac{\mathrm{d}^{2}x}{\mathrm{d}t^{2}}.LEARN
In straight-line motion: at rest when v=0v = 0; changes direction when vv changes sign; speed =∣v∣= \lvert v\rvert, the magnitude of the velocity.LEARN
In straight-line motion, an object speeds up when vv and aa have the same sign and slows down when their signs differ.LEARN
displacement=∫t1t2v(t) dt\text{displacement} = \int_{t_{1}}^{t_{2}} v(t)\,\mathrm{d}tBOOKLET
distance=∫t1t2∣v(t)∣ dt\text{distance} = \int_{t_{1}}^{t_{2}} \lvert v(t)\rvert\,\mathrm{d}tBOOKLET
Distance by hand: split the interval at each time vv changes sign, integrate vv over each piece, and add the sizes.LEARN

5.14 ◆Setting up and separating differential equations

"Proportional to" means "equals a constant kk times". For a quantity that decreases, write −k-k with k>0k > 0.LEARN
Separating: dydx=f(x) g(y)\frac{\mathrm{d}y}{\mathrm{d}x} = f(x)\,g(y), with g(y)≠0g(y) \neq 0, gives ∫1g(y) dy=∫f(x) dx\int \frac{1}{g(y)}\,\mathrm{d}y = \int f(x)\,\mathrm{d}x. Add one + C+\,C as soon as you integrate.LEARN
The general solution of a first-order differential equation has one arbitrary constant. A known value, such as the value at t=0t = 0, fixes it and gives the particular solution.LEARN
dydx=ky\frac{\mathrm{d}y}{\mathrm{d}x} = ky has general solution y=Aekxy = A\mathrm{e}^{kx}, where AA is the value of yy at x=0x = 0.LEARN
ef(x)+c=Aef(x)\mathrm{e}^{f(x) + c} = A\mathrm{e}^{f(x)}: the constant multiplies. Drop the modulus in ln⁡∣y∣\ln|y| only when the context fixes the sign of yy.LEARN

5.15 ◆Slope fields

The slope field of dydx=f(x,y)\frac{\mathrm{d}y}{\mathrm{d}x} = f(x, y) has a short segment at each grid point (a,b)(a, b) with gradient f(a,b)f(a, b). A solution curve is tangent to every segment it passes through.LEARN
In the slope field of dydx=f(x,y)\frac{\mathrm{d}y}{\mathrm{d}x} = f(x, y): if ff contains xx only, the segments in each vertical column are parallel. If ff contains yy only, the segments along each horizontal row are parallel.LEARN
The segments of dydx=f(x,y)\frac{\mathrm{d}y}{\mathrm{d}x} = f(x, y) are horizontal where f(x,y)=0f(x, y) = 0. A solution curve has a stationary point exactly where it meets this curve.LEARN
To sketch a solution on a slope field: go through the given point, follow the segments in both directions across the diagram, and put the turning points where the segments are flat.LEARN
For dydt=f(y)\frac{\mathrm{d}y}{\mathrm{d}t} = f(y) with f(c)=0f(c) = 0, the line y=cy = c is a solution. If the segments just below it rise and those just above it fall, solutions that start close to cc move towards cc as tt increases.LEARN

5.16 ◆Euler's method

yn+1=yn+h×f(xn,yn),xn+1=xn+hy_{n+1} = y_{n} + h \times f(x_{n}, y_{n}), \quad x_{n+1} = x_{n} + hBOOKLET
xn+1=xn+hf1(xn,yn,tn),yn+1=yn+hf2(xn,yn,tn),tn+1=tn+hx_{n+1} = x_{n} + h f_1(x_{n}, y_{n}, t_{n}), \quad y_{n+1} = y_{n} + h f_2(x_{n}, y_{n}, t_{n}), \quad t_{n+1} = t_{n} + hBOOKLET
Euler's method gives an estimate, not the exact solution. A smaller step length means more steps and usually a closer estimate.LEARN
Number of steps =end value−start valueh= \frac{\text{end value} - \text{start value}}{h}. Each row uses only the row above, and in a coupled system both new values come from the old row.LEARN
Spreadsheet: put x0x_{0} and y0y_{0} in row 2, write row 3 as x+hx + h and y+h×f(x,y)y + h \times f(x, y) using row 2's cells, then fill down.LEARN

5.17 ◆Phase portraits for coupled equations

For dxdt=ax+by\frac{\mathrm{d}x}{\mathrm{d}t} = ax + by, dydt=cx+dy\frac{\mathrm{d}y}{\mathrm{d}t} = cx + dy with non-zero eigenvalues, the only equilibrium point is the origin. A trajectory's gradient is dydx=dy/dtdx/dt\frac{\mathrm{d}y}{\mathrm{d}x} = \frac{\mathrm{d}y/\mathrm{d}t}{\mathrm{d}x/\mathrm{d}t}.LEARN
x=Aeλ1tp1+Beλ2tp2(real distinct eigenvalues)\boldsymbol{x} = A\mathrm{e}^{\lambda_{1}t}\boldsymbol{p}_{1} + B\mathrm{e}^{\lambda_{2}t}\boldsymbol{p}_{2}\quad\text{(real distinct eigenvalues)}BOOKLET
Real eigenvalues: each eigenvector line through the origin holds straight-line trajectories, moving in for a negative eigenvalue and out for a positive one.LEARN
For dxdt=ax+by\frac{\mathrm{d}x}{\mathrm{d}t} = ax + by, dydt=cx+dy\frac{\mathrm{d}y}{\mathrm{d}t} = cx + dy, the origin is a stable equilibrium, with every solution heading into it, when both eigenvalues are negative or have a negative real part. When both are positive or have a positive real part, every solution except the origin itself heads away.LEARN
Real eigenvalues of opposite signs: the origin is a saddle point. Trajectories come in alongside the negative eigenvalue's line and leave alongside the positive one's.LEARN
Complex eigenvalues α±βi\alpha \pm \beta\mathrm{i} (α,β≠0\alpha, \beta \neq 0) give spirals. Imaginary eigenvalues ±βi\pm\beta\mathrm{i} (β≠0\beta \neq 0) give closed circles or ellipses. In both cases the motion is anticlockwise if dydt>0\frac{\mathrm{d}y}{\mathrm{d}t} > 0 at (1,0)(1, 0), clockwise if dydt<0\frac{\mathrm{d}y}{\mathrm{d}t} < 0.LEARN

5.18 ◆Second-order differential equations

To solve d2xdt2=f(x,dxdt,t)\frac{\mathrm{d}^{2}x}{\mathrm{d}t^{2}} = f\left(x, \frac{\mathrm{d}x}{\mathrm{d}t}, t\right), let y=dxdty = \frac{\mathrm{d}x}{\mathrm{d}t}: then dxdt=y\frac{\mathrm{d}x}{\mathrm{d}t} = y and dydt=f(x,y,t)\frac{\mathrm{d}y}{\mathrm{d}t} = f(x, y, t). You need two initial conditions, xx and dxdt\frac{\mathrm{d}x}{\mathrm{d}t} at the start.LEARN
Euler's method for a second-order equation, with y=dxdty = \frac{\mathrm{d}x}{\mathrm{d}t}: xn+1=xn+hynx_{n+1} = x_{n} + hy_{n} and yn+1=yn+h×f(xn,yn,tn)y_{n+1} = y_{n} + h \times f(x_{n}, y_{n}, t_{n}), both from the old row. Keep full values until the end.LEARN
For d2xdt2+adxdt+bx=0\frac{\mathrm{d}^{2}x}{\mathrm{d}t^{2}} + a\frac{\mathrm{d}x}{\mathrm{d}t} + bx = 0: the matrix is (01−b−a)\begin{pmatrix}0 & 1\\ -b & -a\end{pmatrix}, the eigenvalues solve λ2+aλ+b=0\lambda^{2} + a\lambda + b = 0, and (1λ)\begin{pmatrix}1\\ \lambda\end{pmatrix} is an eigenvector for each eigenvalue λ\lambda.LEARN
x=Aeλ1tp1+Beλ2tp2(real distinct eigenvalues)\boldsymbol{x} = A\mathrm{e}^{\lambda_{1}t}\boldsymbol{p}_{1} + B\mathrm{e}^{\lambda_{2}t}\boldsymbol{p}_{2}\quad\text{(real distinct eigenvalues)}BOOKLET
For d2xdt2+adxdt+bx=0\frac{\mathrm{d}^{2}x}{\mathrm{d}t^{2}} + a\frac{\mathrm{d}x}{\mathrm{d}t} + bx = 0 with real distinct eigenvalues: x=Aeλ1t+Beλ2tx = A\mathrm{e}^{\lambda_{1}t} + B\mathrm{e}^{\lambda_{2}t}. Find AA and BB from xx and dxdt\frac{\mathrm{d}x}{\mathrm{d}t} at t=0t = 0.LEARN
Phase portrait of xx against y=dxdty = \frac{\mathrm{d}x}{\mathrm{d}t} for d2xdt2+adxdt+bx=0\frac{\mathrm{d}^{2}x}{\mathrm{d}t^{2}} + a\frac{\mathrm{d}x}{\mathrm{d}t} + bx = 0: real, distinct, negative eigenvalues mean a return to rest without oscillating; complex with negative real part, oscillations that die away (a clockwise spiral in); imaginary, oscillations for ever (a clockwise ellipse).LEARN

Which method when

'Find f′(x)f'(x)' for a sum of terms axnax^{n}, nn an integer → power rule term by term, after rewriting axn\frac{a}{x^{n}} as ax−nax^{-n} and expanding brackets; a model with ex\mathrm{e}^{x} → GDC numerical derivative, which gives only the value at one point (5.3, 5.1)LEARN
'Find the values of xx for which ff is increasing (decreasing)' → solve f′(x)=0f'(x) = 0, note any xx where ff or f′f' is undefined, find the sign of f′(x)f'(x) on each piece; answer with inequalities inside the domain, not end points alone (5.2)LEARN
'Equation of the tangent (normal) at x=ax = a' → point (a,f(a))(a, f(a)), gradient f′(a)f'(a), then y−y1=m(x−x1)y - y_{1} = m(x - x_{1}); the normal's gradient is −1f′(a)-\frac{1}{f'(a)} when f′(a)≠0f'(a) \neq 0, and the normal is x=ax = a when f′(a)=0f'(a) = 0 (5.4)LEARN
'Coordinates of the local maximum (minimum)' → solve f′(x)=0f'(x) = 0 (GDC: the zeros of y=f′(x)y = f'(x), or the polynomial solver when f′(x)f'(x) is a polynomial), yy from f(x)f(x); classify by the sign change of f′(x)f'(x); for the greatest value on a≤x≤ba \le x \le b of a graph with no breaks, also compare f(a)f(a) and f(b)f(b) (5.6)LEARN
'Maximise the volume', 'least cost', a fixed area or amount of material → use the constraint to get one variable, solve f′(x)=0f'(x) = 0, justify with the sign change of f′(x)f'(x), compare with the end points of the domain; a whole-number variable: compare the whole numbers either side (5.7)LEARN
Given dydx\frac{\mathrm{d}y}{\mathrm{d}x} and a point, 'find yy in terms of xx' → integrate term by term with + C+\,C (n≠−1n \neq -1), substitute the point into the integrated function (not the derivative), write the full function; one point fixes CC only on a domain with no breaks (5.5)LEARN
'Total change' from a rate, or 'area under the curve' where f(x)≥0f(x) \ge 0 for a≤x≤ba \le x \le b → write ∫abf(x) dx\int_{a}^{b} f(x)\,\mathrm{d}x, then evaluate it on the GDC; for a region enclosed with the xx-axis, the limits are the curve's zeros, kept in full (5.5)LEARN
A table at equal intervals, or 'use the trapezoidal rule' → 12h((y0+yn)+2(y1+⋯+yn−1))\frac{1}{2}h\left((y_{0} + y_{n}) + 2(y_{1} + \dots + y_{n-1})\right) with nn intervals, n+1n + 1 values and h=b−anh = \frac{b - a}{n}; concave up across the whole interval overestimates, concave down across the whole interval underestimates (5.8)LEARN

◆Which method when

A rate linked to another quantity's rate, 'how fast is the area increasing when r=3r = 3?' → related rates: an equation true at every moment, differentiate, write the chain rule such as dAdt=dAdr×drdt\frac{\mathrm{d}A}{\mathrm{d}t} = \frac{\mathrm{d}A}{\mathrm{d}r} \times \frac{\mathrm{d}r}{\mathrm{d}t}, then substitute the values at the instant; a decreasing quantity has a negative rate (5.9)LEARN
'Justify that it is a maximum' → second derivative test at a stationary point: f′′(a)<0f''(a) < 0 maximum, f′′(a)>0f''(a) > 0 minimum, and if f′′(a)=0f''(a) = 0 use the sign of f′f' either side; 'when is it increasing fastest?' → the maximum of f′f': a point of inflexion where f′′f'' changes from positive to negative, or the maximum of y=f′(x)y = f'(x) on the GDC; on a closed domain also compare the end points (5.10)LEARN
'Find the exact value' of an integral → antiderivative by hand: a linear inside ax+bax + b → integrate as if it were xx and divide by aa; a constant times g′(x) f(g(x))g'(x)\,f(g(x)), including g′(x)g(x)→ln⁡∣g(x)∣\frac{g'(x)}{g(x)} \to \ln|g(x)| → inspection or substitution u=g(x)u = g(x), changing the limits to uu-values; the GDC value is only a check (5.11)LEARN
Area with part of the curve below the xx-axis → split at the crossings and add the sizes, or ∫ab∣y∣ dx\int_{a}^{b} |y|\,\mathrm{d}x on the GDC; region against the yy-axis → ∫∣x∣ dy\int |x|\,\mathrm{d}y with yy-limits; a full turn about the xx-axis → ∫πy2 dx\int \pi y^{2}\,\mathrm{d}x, about the yy-axis → ∫πx2 dy\int \pi x^{2}\,\mathrm{d}y with yy-limits (5.12)LEARN
Straight-line motion → differentiate ss for vv and aa; vv given in terms of ss → a=vdvdsa = v\frac{\mathrm{d}v}{\mathrm{d}s}; 'displacement' from t1t_{1} to t2t_{2} → ∫t1t2v dt\int_{t_{1}}^{t_{2}} v\,\mathrm{d}t; 'distance travelled' → ∫t1t2∣v∣ dt\int_{t_{1}}^{t_{2}} |v|\,\mathrm{d}t on the GDC; speeding up when vv and aa have the same sign (5.13)LEARN
'The rate is proportional to…' → dydt=k×(…)\frac{\mathrm{d}y}{\mathrm{d}t} = k \times (\ldots), with −k-k and k>0k > 0 for a decrease; right side of the form f(x) g(y)f(x)\,g(y) → separate the variables (dividing by g(y)≠0g(y) \neq 0), add one + C+\,C as you integrate, then use the known value; dydx=ky\frac{\mathrm{d}y}{\mathrm{d}x} = ky → y=Aekxy = A\mathrm{e}^{kx} (5.14)LEARN
A differential equation you can't separate → 'sketch the solution through a point' or 'describe the long-term behaviour': slope field, follow the segments both ways, turning points where f(x,y)=0f(x, y) = 0; 'use Euler's method to estimate': yn+1=yn+hf(xn,yn)y_{n+1} = y_{n} + h f(x_{n}, y_{n}) on the GDC or a spreadsheet, end−starth\frac{\text{end} - \text{start}}{h} steps, and for a coupled pair both new values from the old row (5.15, 5.16)LEARN
Linear coupled pair dxdt=ax+by\frac{\mathrm{d}x}{\mathrm{d}t} = ax + by, dydt=cx+dy\frac{\mathrm{d}y}{\mathrm{d}t} = cx + dy, 'phase portrait' or 'long-term behaviour' → eigenvalues of the matrix: real distinct → x=Aeλ1tp1+Beλ2tp2\boldsymbol{x} = A\mathrm{e}^{\lambda_{1}t}\boldsymbol{p}_{1} + B\mathrm{e}^{\lambda_{2}t}\boldsymbol{p}_{2}, both negative → stable, opposite signs → saddle point; complex with real part ≠0\neq 0 → spiral, inwards if the real part is negative; purely imaginary → closed ellipses; turning direction from the sign of dydt\frac{\mathrm{d}y}{\mathrm{d}t} at (1,0)(1, 0) (5.17)LEARN
Second-order d2xdt2+adxdt+bx=0\frac{\mathrm{d}^{2}x}{\mathrm{d}t^{2}} + a\frac{\mathrm{d}x}{\mathrm{d}t} + bx = 0 → let y=dxdty = \frac{\mathrm{d}x}{\mathrm{d}t}, matrix (01−b−a)\begin{pmatrix}0 & 1\\ -b & -a\end{pmatrix}, eigenvalues from λ2+aλ+b=0\lambda^{2} + a\lambda + b = 0; real distinct → x=Aeλ1t+Beλ2tx = A\mathrm{e}^{\lambda_{1}t} + B\mathrm{e}^{\lambda_{2}t} with AA and BB from both xx and dxdt\frac{\mathrm{d}x}{\mathrm{d}t} at t=0t = 0; any other second-order equation → Euler's method on the coupled form (5.18)LEARN
Follows the IB Maths AI guide (first assessment 2021). Not affiliated with the IB.