Sign in to save your mistake log, flashcard schedule, and streaks across every device — and to use the AI tutor and question generator.
Also unlocks Google Forms export for generated quizzes.
or use email
🏫 Classes
Create a class to get a code you can share with students, or join one with a code your teacher gave you. You can do both.
Join a class
Create a class
🔗 Connect Google Account
IB · International Baccalaureate
IB Mathematics HL
Courses · Analysis & Approaches (AA) · Applications & Interpretation (AI)Exam · Paper 1 (no calc) · Paper 2 (GDC) · Paper 3 (extended investigation)Topics · 5 core topics + HL extensions
i
Overview
About IB Mathematics HL
Analysis and Approaches HL (AA HL) is the most rigorous IB math course — equivalent to a first-year university pure mathematics course. It extends AA SL with complex numbers, proof by induction, vectors in 3D, advanced integration (by parts, substitution, partial fractions), differential equations, and series convergence.
Applications and Interpretation HL (AI HL) extends AI SL with matrices, complex numbers, Poisson distribution, graph theory (networks), and deeper optimization and differential equations.
IB HL Syllabus Map (AA & AI)
Topic
AA HL specifics
AI HL specifics
~Marks
1 — Algebra
Proof by induction, complex numbers
Matrices, eigenvalues
20
2 — Functions
Modulus, complex roots, partial fractions
Additional regression
20
3 — Geometry & Trig
3D vectors, cross product, planes
Graph theory, algorithms
25
4 — Statistics
Unbiased estimators, CIs, hypothesis tests
Poisson, Markov chains
30
5 — Calculus
By parts, DEs, Maclaurin series
Coupled DEs, numerics
40
Exam: Paper 1 (no GDC, 120 min) + Paper 2 (GDC, 120 min) + Paper 3 (extended, 60 min). Internal Assessment 20%.
Every official HL subtopic above links straight to the lesson that covers it — "Core" rows point to the shared SL lesson (same content at HL), everything else to this track's own HL-extension chapters. Click any lesson name to jump there.
Topic 1 HL
Proof by induction, complex numbers, partial fractions, sum of series, system of linear equations.
Topic 2 HL
Rational functions, polynomial long division, modulus function, odd/even, complex roots.
Topic 3 HL
3D vectors, lines/planes in 3D, cross product, intersections, angles between planes.
Topic 4 HL
Poisson distribution (AI), unbiased estimators, confidence intervals, hypothesis testing.
Exam (both courses): Paper 1 (no GDC, 120 min, 110 marks) + Paper 2 (GDC, 120 min, 110 marks) + Paper 3 (60 min, 55 marks — investigative, GDC required).
⚠ IB Mathematics HL Exam Traps & Command Terms
Accuracy Standard: Non-exact numerical answers must be given to exactly 3 significant figures unless specified otherwise. Intermediate rounding errors incur an automatic accuracy mark penalty (A1).
Command Terms:'Show that' means the answer is already provided; every single intermediate algebraic step must be justified without skipping. 'Hence' means you must use the preceding part's result; alternate methods receive no credit.
Paper 1 vs Paper 2: Paper 1 prohibits any calculator; you must show exact values involving fractions, surds, e, and π. Paper 2 and 3 require GDC fluency (graphing, matrix solvers, numerical integration, and statistics distributions).
IB Mathematics HL assessment component weighting: Paper 1 (30%), Paper 2 (30%), Paper 3 (20%), and Internal Assessment (20%).
A student receives the following scaled marks on IB AA HL assessment: Paper 1: 85/110, Paper 2: 92/110, Paper 3: 44/55, and IA: 16/20. Calculate the student's overall weighted percentage.
1. Compute the component percentages: Paper 1: 11085≈77.27% Paper 2: 11092≈83.64% Paper 3: 5544=80.00% IA: 2016=80.00%
HL Topic 1 extends SL algebra with proof by induction and complex numbers — the foundation for much of higher mathematics.
● Proof by mathematical induction
Structure: (1) Base case: verify for n=1 (or initial n). (2) Inductive hypothesis: assume true for n=k. (3) Inductive step: prove true for n=k+1. (4) Conclude.
● Complex numbers
z=a+bi,i2=−1,Re(z)=a,Im(z)=b.∣z∣=a2+b2,arg(z)=arctan(b/a) (adjusted for quadrant).zˉ=a−bi,zzˉ=∣z∣2,z1=∣z∣2zˉ.
⊙ Explorer · Complex numbers on the Argand planeinteractive
z, its conjugate, modulus, and argument
quantity
value
|z|5.000
arg(z)53.13°
z̄ (conjugate)3.00 − 4.00i
z = a + bi (indigo dot) and its conjugate z̄ = a − bi (amber, dashed segment) are mirror images across the real axis — exactly what negating the imaginary part does geometrically. |z| is the dot's distance from the origin; arg(z) is the angle the segment from the origin to z makes with the positive real axis, drawn as an arc. Drag a or b negative to move z into another quadrant and watch arg(z) jump to the correctly-adjusted value, not just arctan(b/a) taken literally.
n-th roots of unity: solutions to zn=1 are e2πik/n for k=0,1,…,n−1.
⊙ Explorer · De Moivre's theorem on the unit circleinteractive
The n-th roots of unity: angle and coordinates of each
k
angle
cos
sin
zⁿcis(90.0°)
root spacing (360°/n)120.0°
rⁿ = |zⁿ|1.000
z = r·cis θ (indigo) and zⁿ = rⁿ·cis(nθ) (amber, dashed) — this is de Moivre's theorem in full: raising z to the nth power multiplies the angle by n AND raises the modulus to the nth power. The gray rings are the n-th roots of unity (always on the unit circle, r = 1), k = 0, …, n−1, evenly spaced 360°/n apart starting at angle 0°. Set r = 1 and drag θ to a multiple of that spacing to land zⁿ exactly on z = 1, the k = 0 root. Push r above or below 1 and watch zⁿ leave the unit circle — outward for r > 1, inward for r < 1 — since |zⁿ| = rⁿ grows or shrinks with n, unlike the roots of unity themselves.
A system of up to three linear equations in three unknowns can be solved by elimination: combine pairs of equations to cancel one variable at a time, reducing to a smaller system, until a single equation in one unknown remains — then substitute back up.
Geometrically, each equation ax+by+cz=d is a plane in 3D (Chapter 3 of this track studies these directly). Solving the system asks how three planes intersect:
Unique solution — the three planes meet at exactly one point (the generic case).
No solution — elimination produces a contradiction like 0=k for k=0; geometrically, no point lies on all three planes at once (e.g. two are parallel, or the three form a triangular "tube" with no common point).
Infinitely many solutions — elimination produces an identity like 0=0, leaving a free parameter; geometrically, the planes share a common line (or are all the same plane).
Worked Example 1.4 — Solving a 3×3 system by elimination
Problem.Solve: x+y+z=6, 2x−y+z=3, x+2y−z=2.
Label the equations (1), (2), (3). Add (1) and (3) to eliminate z: (x+y+z)+(x+2y−z)=6+2⇒2x+3y=8 — call this (4).
Subtract (2) from (1) to eliminate z again: (x+y+z)−(2x−y+z)=6−3⇒−x+2y=3⇒x=2y−3 — call this (5).
Substitute (5) into (4): 2(2y−3)+3y=8⇒4y−6+3y=8⇒7y=14⇒y=2.
Then x=2(2)−3=1, and from (1): 1+2+z=6⇒z=3.
Check in (2): 2(1)−2+3=3. ✓ Check in (3): 1+2(2)−3=2. ✓
Solution: (x,y,z)=(1,2,3) — the unique point where all three planes meet.
Worked Example 1.1 — Proof by induction
Problem.Prove by induction: r=1∑nr2=6n(n+1)(2n+1).
Base (n=1): LHS=1. RHS=61⋅2⋅3=1.✓
Assumer=1∑kr2=6k(k+1)(2k+1).
For n=k+1:r=1∑k+1r2=6k(k+1)(2k+1)+(k+1)2=(k+1)[6k(2k+1)+(k+1)]
Step: Assume true for n=k. Then ∑r=1k+1r=2k(k+1)+(k+1)=2(k+1)(k+2). ✓
Practice: Find the four fourth roots of 16i in polar form.
16i=16eiπ/2. Roots: 2ei(π/8+kπ/2) for k=0,1,2,3.
Worked Example · A system with infinitely many solutions
Solve: x+z=2, y=1, x+y+z=3.
Solution: The third equation is exactly the sum of the first two ((x+z)+y=2+1=3) — it carries no new information.
From the first two: z=2−x, y=1, with x free. Infinitely many solutions:(x,y,z)=(t,1,2−t) for any t∈R — a whole line, exactly the "shared line" case above.
Polar form of complex number z = 1 + i√3 = 2·cis(π/3) plotted in the complex plane with modulus r = 2 and argument θ = π/3.
Induction Conclusion Statement: IB mark schemes reserve an explicit final mark (R1) for the full concluding sentence: 'Since true for n=1, and if true for n=k then true for n=k+1, by mathematical induction the statement is true for all n ∈ ℤ⁺.' Omitting this loses the mark.
Principal Argument Interval: In IB AA HL, the argument must satisfy −π<Arg(z)≤π. Angles like 300∘ or 35π must be rewritten as −3π.
Partial Fraction Degree Check: For improper rational fractions where degree of numerator ≥ degree of denominator, you must perform polynomial division before setting up partial fractions.
HL functions extend SL content with the modulus function, rational functions with oblique asymptotes, and complex roots of polynomials.
● Modulus (absolute value) function
∣f(x)∣: reflect negative part above x-axis.f(∣x∣): reflect positive part across y-axis.
Equations ∣f(x)∣=k give two cases: f(x)=k and f(x)=−k. Inequalities ∣f(x)∣<k give −k<f(x)<k.
⊙ Explorer · y = f(x) vs y = |f(x)|, and solving |f(x)| = kinteractive
Sampled points on f(x) and the currently displayed curve
x
f(x)
displayed y
real solutions in this window1
f(x) = 0.5x³ − 2x (faint grey when |f(x)| is shown, so you can see exactly which parts got reflected). Tick the checkbox to fold the curve's negative dips up above the x-axis, matching |f(x)|'s definition above. The amber line y = k marks where you'd read off solutions to f(x) = k (or |f(x)| = k) — count its crossings with the displayed curve; that count is exactly what the readout reports, live.
● Factor/remainder theorem (polynomials)
Complex conjugate root theorem: if p(x) has real coefficients and a+bi is a root, then a−bi is also a root.
Relation between roots and coefficients for az3+bz2+cz+d=0 with roots α,β,γ:
α+β+γ=−b/a,αβ+βγ+γα=c/a,αβγ=−d/a.
Worked Example 2.1
Problem.Sketch y=∣2x−4∣ and solve ∣2x−4∣=x+1.
V-shape with vertex at (2,0).
Case 1: 2x−4=x+1⇒x=5. Check: ∣6∣=6=6.✓
Case 2: 2x−4=−(x+1)⇒3x=3⇒x=1. Check: ∣−2∣=2=2.✓
Worked Example 2.2 — Complex roots
Problem.p(x)=x3−3x2+7x−5 has root x=1. Find all roots.
Factor out (x−1): p(x)=(x−1)(x2−2x+5).
Quadratic: x=22±4−20=1±2i.
Roots: x=1,1+2i,1−2i.
Modulus function
∣f(x)∣ reflects negative output above x-axis
Odd/even symmetry
Odd: f(−x)=−f(x); Even: f(−x)=f(x)
Practice Problem
p(x)=x4−2x3+5x2−8x+4 has x=1 as a double root. Find all roots.
Show solution
(x−1)2 divides p(x). Divide: p(x)=(x−1)2(x2+4).
x2+4=0⇒x=±2i. Roots: 1 (double), ±2i.
Worked Example · Rational function sketch
Sketch f(x)=x2−42x2−x−6, stating all asymptotes and intercepts.
Solution: Factor: f(x)=(x+2)(x−2)(2x+3)(x−2)=x+22x+3 for x=2.
Case 2: 2x−3=−(x+1)⟹3x=2⟹x=32. Check: ∣−35∣=35=32+1. ✓
Graph of modulus function f(x) = |2x - 3| showing vertex cusp at x = 1.5 and reflection of negative values across the x-axis.
⚠ Topic 2 Exam Traps: Modulus & Inverse Functions
Squaring Modulus Equations: While squaring both sides of ∣ax+b∣=∣cx+d∣ is convenient, squaring when one side is variable without modulus (e.g. ∣2x−3∣=x−1) can introduce extraneous roots. Always check candidates in the un-squared equation.
Self-Inverse Functions: For rational functions of the form f(x)=cx−aax+b, f(x)=f−1(x). Recognize this symmetry immediately to save algebraic time.
Domain Restrictions on Inverses: For quadratic functions, an inverse function only exists if the domain is restricted to one side of the vertex (e.g. x≥h). Stating f−1 without stating its restricted domain x≥k loses marks.
Practice: Find the range of f(x)=x2+1x.
Let y=x2+1x. Then yx2−x+y=0. For real x: Δ=1−4y2≥0⟹y∈[−21,21].
📝 Chapter Quiz
3
Vectors & 3D Geometry
IB Mathematics HL P1, P2 & P3AA HLAI HL
HL extends vectors to three dimensions, covering lines and planes in space, the cross product, and intersection problems — foundational for physics and engineering.
The compound angle formulae extend SL's double-angle identities (Chapter 3 of the SL track) to a sum or difference of any two angles: sin(A±B)=sinAcosB±cosAsinB,cos(A±B)=cosAcosB∓sinAsinB,tan(A±B)=1∓tanAtanBtanA±tanB. Setting A=B=θ in the first two recovers exactly the SL double-angle identities sin2θ=2sinθcosθ and cos2θ=cos2θ−sin2θ — those are this chapter's special case, not a separate fact.
The reciprocal trigonometric functions are secθ=cosθ1,cscθ=sinθ1,cotθ=tanθ1=sinθcosθ. Dividing the Pythagorean identity sin2θ+cos2θ=1 through by cos2θ (then by sin2θ) gives two more standard identities: 1+tan2θ=sec2θ,1+cot2θ=csc2θ.
Worked Example · An exact value via compound angles
Find the exact value of sin75° using a compound angle formula.
Solution: Write 75°=45°+30°, two angles with known exact sine/cosine values: sin75°=sin(45°+30°)=sin45°cos30°+cos45°sin30°.
=22⋅23+22⋅21=46+42=46+2.
Worked Example · Proving a reciprocal identity
Prove 1+cot2θ=csc2θ.
Solution: Start from sin2θ+cos2θ=1 and divide every term by sin2θ (valid whenever sinθ=0):
Drag to rotate, scroll to zoom · loads a small 3D library on first use
The two vectors and their cross product
vector
⟨x, y, z⟩
magnitude
u × v⟨7, 10, -9⟩
|u × v|15.166
u=⟨1,2,3⟩ fixed (blue), v=⟨v1,−1,v3⟩ (amber, both components adjustable). u×v (green) always stays perpendicular to both — drag either slider and confirm it never tilts away from that right angle.
Worked Example 3.2 — Intersection of planes
Problem.Find the line of intersection of planes π1:x+2y−z=3 and π2:2x−y+3z=1.
Direction of line: n1×n2=⟨1,2,−1⟩×⟨2,−1,3⟩=⟨6−1,−3−2,−1−4⟩=⟨5,−5,−5⟩∝⟨1,−1,−1⟩.
Find a point: set z=0: x+2y=3,2x−y=1. Solve: x=1,y=1. Point (1,1,0).
Line: r=110+t1−1−1.
Line in 3D
r=a+td
Plane equation
r⋅n=a⋅n or ax+by+cz=d
Practice Problem
Find the equation of the plane through A(1,2,3), B(4,0,1), C(2,3,0).
Show solution
AB=⟨3,−2,−2⟩,AC=⟨1,1,−3⟩.
n=AB×AC=⟨8,7,5⟩.
Plane: 8(x−1)+7(y−2)+5(z−3)=0⇒8x+7y+5z=37.
Worked Example · Angle between two planes
Find the angle between planes π1:2x−y+2z=5 and π2:x+2y−2z=3.
Solution:n1=(2,−1,2), n2=(1,2,−2).
cosθ=∣n1∣∣n2∣∣n1⋅n2∣=3⋅3∣2−2−4∣=94⟹θ≈63.6°
Worked Example · Line–plane intersection
Line: r=(1,0,2)+t(1,1,−1). Plane: x+2y+z=4. Find the intersection.
Vector geometry in 3D: vectors AB and AC defining a plane with normal vector n = AB × AC perpendicular to both sides.
⚠ Topic 3 Exam Traps: 3D Vector Lines & Planes
Cross Product Anti-Commutativity:a×b=−(b×a). Pay strict attention to order when computing normal vectors.
Angle Between Line and Plane: The angle ϕ between line direction d and plane normal n gives cosϕ=∣d∣∣n∣∣d⋅n∣. But the angle θ between the line and the plane itself is 90∘−ϕ, so sinθ=∣d∣∣n∣∣d⋅n∣. Using cosine for the line-plane angle is a classic trap.
Skew vs Intersecting Lines: In 3D, two lines that are not parallel might not intersect. You must test whether there is a simultaneous solution for parameter values λ and μ; if not, the lines are skew.
Practice: Find the distance from point P(3,1,−2) to plane 2x−2y+z=5.
d=4+4+1∣2(3)−2(1)+(−2)−5∣=3∣6−2−2−5∣=31
📝 Chapter Quiz
4
Statistics & Probability (HL)
IB Mathematics HL P1, P2 & P3AA HLAI HL
HL extends SL statistics with unbiased estimators, confidence intervals, hypothesis testing, and (AI HL) the Poisson distribution.
● Unbiased estimators
μ^=xˉ=n∑xi(unbiased for μ).S2=n−1∑(xi−xˉ)2(unbiased for σ2;use sn−1 on GDC).
Common critical values: 90% CI → z=1.645; 95% CI → z=1.96; 99% CI → z=2.576.
⊙ Explorer · Confidence interval for the meaninteractive
Sample statistics and the resulting interval
quantity
value
margin of error2.058
confidence interval(50.24, 54.36)
This explorer uses the z-based formula xˉ±zα/2⋅s/n throughout (a good approximation once n is reasonably large) rather than switching to the t-interval Worked Example 4.1 uses for its exact small-sample answer — expect the shaded interval here to be a touch narrower than that example's t-based one at the same 95%. The curve is the sampling distribution of the mean, width s/n — shrink it by raising n, or widen the shaded region by demanding a higher confidence level; both change the interval for the reason the formula says they should.
● Poisson distribution (AI HL)
X∼Po(λ):P(X=k)=k!e−λλk,μ=σ2=λ.
Conditions: events occur independently at a constant average rate λ per unit interval.
Poisson as approximation to Binomial: when n is large, p is small, and np=λ is moderate.
Worked Example 4.1 — CI
Problem.A sample of n=64 gives xˉ=52.3 and s=8.4. Construct a 95% CI for μ.
Use t-interval (σ unknown): t0.025,63≈2.00 (from tables or GDC).
Default λ=3, k=5 is the toll-booth rate and outcome from Worked Example 4.2 — the amber bar is exactly that example's answer. Drag k to move the highlighted bar and read off any P(X=k) directly; drag λ and watch the whole distribution's peak shift and spread out (mean = variance = λ, so a larger rate means a wider, flatter-looking bell).
Unbiased estimator
s2=n−1∑(xi−xˉ)2 estimates σ2
Hypothesis test
State H0,H1; compute test stat; compare p-value to α
Practice Problem
A machine produces defective items at 2 per hour (Poisson). Find P(exactly 3 defective) in a 2-hour period.
Show solution
λ=4.P(X=3)=3!e−4⋅43=664e−4≈0.1954.
Worked Example · Hypothesis test for population mean
A sample of 25 gives xˉ=48.2, s=6. Test H0:μ=50 vs H1:μ=50 at 5% significance.
Solution:t=6/2548.2−50=1.2−1.8=−1.5.
Critical value t24,0.025≈2.064. Since ∣−1.5∣<2.064, do not reject H0.
Worked Example · Confidence interval
Construct a 95% CI for the mean of a normal population with known σ=8 given n=36, xˉ=62.
Sample Variance Divisor: The unbiased estimator of population variance uses denominator n−1: sn−12=n−11∑(xi−xˉ)2. The calculator σx uses n; always use sx (the n−1 version) in IB HL inference.
Poisson Parameter Scaling: If events occur at rate λ per hour, the rate for 30 minutes is 0.5λ, and for 2 hours is 2λ. Scale λ before computing probabilities.
Hypothesis Test Conclusion: Never write 'accept H0'. Write 'There is insufficient evidence at the 5% level to reject H0', followed by context in terms of the problem.
Practice: Two independent samples: n1=10,xˉ1=20,s12=4 and n2=15,xˉ2=18,s22=5. Test H0:μ1=μ2 at 10% using a t-test. State your conclusion.
Pooled sp2=239(4)+14(5)=23106≈4.61. t=4.61(1/10+1/15)2≈0.7842≈2.55. Critical value t23,0.05≈1.714. Reject H0.
📝 Chapter Quiz
5
Calculus (HL)
IB Mathematics HL P1, P2 & P3AA HLAI HL
HL calculus covers the rigorous theoretical foundations of limits and continuity, advanced integration techniques, differential equations, and Maclaurin series — bridging secondary school and undergraduate pure analysis.
● The 3-Part Definition of Continuity at a Point & Complete Taxonomy of Graph Cases
A function f is continuous at x=c if and only if all three conditions of the Continuity Checklist are satisfied:
Condition 1:f(c) is defined (meaning c is in the domain of f).
Condition 2:limx→cf(x) exists (meaning left and right limits are finite and agree: limx→c−f(x)=limx→c+f(x)=L).
Condition 3:limx→cf(x)=f(c) (the limiting trend matches the actual output).
Failure of any condition produces a specific discontinuity geometry on the Cartesian plane:
Case 1 · Continuous: All 3 conditions hold. The curve passes unbroken through (c,f(c)).
Case 2 · Removable (Hole): Condition 1 fails (undefined at c), but Condition 2 holds (limit L exists). The graph has an open ring at (c,L).
Case 3 · Removable (Displaced Point): Conditions 1 and 2 hold, but Condition 3 fails (f(c)=L). An open ring sits at (c,L) while a solid dot sits at (c,f(c)).
Case 4 · Jump Discontinuity: Condition 2 fails because one-sided limits are finite but unequal (limx→c−f(x)=limx→c+f(x)). A step break occurs in the graph.
Case 5 · Infinite Discontinuity (Vertical Asymptote): Condition 2 fails because one or both one-sided limits grow without bound to ±∞.
Case 6 · Essential / Oscillating Discontinuity: Condition 2 fails because function values oscillate infinitely rapidly between bounds as x approaches c.
⊙ Explorer · All cases of graphs of continuityinteractive
Continuity test checklist and values across sampled inputs
x
f(x)
Left lim x→1⁻
Right lim x→1⁺
Continuity evaluation
input x0.525
output f(x)1.613
lim x→1⁻1.500
lim x→1⁺1.500
Continuity testContinuous at c = 1 ✔
Switch between all 6 continuity graph cases and move x to inspect how the 3-part continuity checklist (f(c) defined, lim x→c exists, lim = f(c)) succeeds or fails.
● Precise (ε, δ) & (ε, M) Definitions of Limits & Rigorous Analysis Proofs
Precise Definition (Limit at a Point):x→climf(x)=L means: for every ε>0, there exists δ>0 such that 0<∣x−c∣<δ⟹∣f(x)−L∣<ε.
Rigorous Proof (Linear): Prove that limx→3(4x−5)=7. Scratchwork:∣(4x−5)−7∣=4∣x−3∣<ε⟺∣x−3∣<4ε. Proof: Let ε>0. Choose δ=4ε. If 0<∣x−3∣<δ, then ∣(4x−5)−7∣=4∣x−3∣<4(4ε)=ε. Thus limx→3(4x−5)=7. ■
Rigorous Proof (Quadratic with Bounding): Prove that limx→2x2=4. Scratchwork:∣x2−4∣=∣x−2∣∣x+2∣. Restrict δ≤1⟹1<x<3⟹∣x+2∣<5. Then ∣x2−4∣<5∣x−2∣. Set 5∣x−2∣<ε⟺∣x−2∣<5ε. Choose δ=min(1,5ε). Proof: Let ε>0. Set δ=min(1,5ε). If 0<∣x−2∣<δ, then ∣x+2∣<5 and ∣x−2∣<5ε. Thus ∣x2−4∣=∣x−2∣∣x+2∣<(5ε)(5)=ε. Hence limx→2x2=4. ■
Precise Definition (Limit at Infinity):x→∞limf(x)=L means: for every ε>0, there exists M>0 such that x>M⟹∣f(x)−L∣<ε.
Rigorous Proof (Limit at Infinity): Prove that limx→∞x+32x+1=2. Scratchwork:x+32x+1−2=x+32x+1−2x−6=x+35<x5<ε⟺x>ε5. Proof: Let ε>0. Choose M=ε5. For all x>M, we have x>ε5⟹x5<ε. Since x>0, x+35<x5<ε. Therefore x+32x+1−2=x+35<ε. Thus limx→∞x+32x+1=2. ■
⊙ Explorer · Precise (ε, δ) limit definitioninteractive
Epsilon-delta tolerance test across domain interval
x
f(x)
|f(x) − L|
Tolerance status
Max valid δ (ε / 2)0.250
Condition statusValid: curve inside ε-band ✔
For any chosen ε-band (horizontal green corridor), a valid δ must keep the entire curve over (c − δ, c + δ) inside the band. Notice δ ≤ ε / 2 guarantees success for f(x) = 2x − 1.
● Integration techniques
Integration by substitution:∫f(g(x))g′(x)dx=∫f(u)du where u=g(x).
Integration by parts:∫udv=uv−∫vdu. (Choose u by LIATE: Log, Inverse trig, Algebraic, Trig, Exponential.)
Partial fractions: decompose rational integrand, then integrate term by term.
⊙ Explorer · The Maclaurin series for cos xinteractive
cos(x) vs. the current Maclaurin polynomial
x
cos(x)
approximation
f(0.1) exact0.995004
Maclaurin approx.0.995004
n=3 with cos x reproduces Worked Example 5.3 exactly (up to the x4 term). Watch the amber curve hug the blue one near x=0 — and peel away further out. Switch the function to sin x to see the same idea apply to the other standard series above — an odd function needs only odd-power terms, so its approximation improves in different-sized jumps than cos x's.
Constant of Integration in DEs: The constant +C must be introduced at the exact moment of integration: ∫y1dy=∫f(x)dx⟹ln∣y∣=F(x)+C. Exponentiating gives ∣y∣=eF(x)+C=AeF(x). Adding +C at the very end as y=eF(x)+C is invalid.
Integration by Parts Sign Error:∫udv=uv−∫vdu. Watch the minus sign when v itself carries a negative sign (e.g. ∫sinxdx=−cosx).
Maclaurin Validity: The series expansion for ln(1+x)=x−2x2+⋯ is only valid for −1<x≤1.
Practice: Use integration by substitution to find ∫1−x2xdx.
Let u=1−x2, du=−2xdx. Integral =−21∫u−1/2du=−1−x2+C.
📝 Chapter Quiz
AA
AA HL — Additional Topics
IB Mathematics HL — AA P1, P2 & P3AA HL only
These topics are specific to Analysis and Approaches HL.
● Convergence of series
Geometric series converges iff ∣r∣<1:S∞=1−ra.Ratio test: if n→∞limanan+1<1⇒series converges absolutely.
⊙ Explorer · Inverse trig functions and their derivativesinteractive
Sampled points on the chosen function and its derivative
x
f(x)
f′(x)
f(x)0.524
f′(x)1.155
Blue is the chosen inverse trig function; amber dashed is its derivative, from the formulas above. sin⁻¹ and cos⁻¹ are only defined on [−1, 1], so their derivative curves diverge near the ends of that interval (division by √(1−x²) → 0) — sin⁻¹'s derivative shoots up toward +∞ at both ends (it's 1/√(1−x²), always positive), while cos⁻¹'s shoots down toward −∞ instead (it's the negative of the same expression) — switch to tan⁻¹ to see a derivative that stays bounded everywhere instead, since 1/(1+x²) never blows up.
Worked Example — Limit and convergence
Problem.Determine if n=1∑∞2nn converges.
Ratio test: anan+1=2n+1n+1⋅n2n=2nn+1→21<1. Series converges.
Mathematical induction
Base case; assume n=k; prove n=k+1
De Moivre's theorem
(cosθ+isinθ)n=cosnθ+isinnθ
Euler's form
eiθ=cosθ+isinθ
Derivative curve f'(x) = 1/(1+x^2) of arctan(x), showing peak at (0, 1) and symmetry about the y-axis.
⚠ Topic AA Exam Traps: Convergence, Inverse Trig & Induction
Ratio Test Inconclusive: If limn→∞anan+1=1, the ratio test provides no conclusion. The series could converge or diverge. You must employ an alternative method such as the comparison or integral test.
Inverse Trig Range Limits: The principal value range of arcsin(x) is [−2π,2π], but arccos(x) has range [0,π]. Never provide negative output angles for arccos(x).
Induction Communication Marks: IB markschemes strictly penalize induction proofs that skip the concluding sentence: 'Since P(1) is true and P(k)⟹P(k+1), by PMI P(n) is true for all n∈Z+'.
Practice: Prove that cos3θ=4cos3θ−3cosθ using De Moivre's theorem.
(cosθ+isinθ)3=cos3θ+isin3θ. Expand left side using binomial; equate real parts.
Practice: Find the Maclaurin series for ln(1+2x) up to the x3 term, and state its interval of convergence.
Use ln(1+u)=u−2u2+3u3−⋯ with u=2x: ln(1+2x)=2x−2x2+38x3−⋯. Convergence requires −1<2x≤1⟹−21<x≤21.
📝 Chapter Quiz
AI
AI HL — Additional Topics
IB Mathematics HL — AI P1, P2 & P3AI HL only
These topics are specific to Applications and Interpretation HL.
● Matrices
For 2×2 matrix A=(acbd):
detA=ad−bc,A−1=ad−bc1(d−c−ba).
Solving Ax=b: x=A−1b. Eigenvalues: det(A−λI)=0.
● Markov chains
A Markov chain models a system that moves between a fixed set of states, where the probability of the next state depends only on the current one (not the history). If pn is the state-probability (column) vector at step n and T is the transition matrix (column j holds the probabilities of moving from state j), then:
pn+1=Tpn⇒pn=Tnp0.
A steady states satisfies Ts=s — it is an eigenvector of T with eigenvalue 1, normalised so its entries sum to 1. For a well-behaved (regular) chain, pn→s as n→∞, regardless of the starting distribution p0.
⊙ Explorer · Markov chain — approach to steady stateinteractive
P(state A) at selected steps
step n
P(A)
steady state πA0.571
P(A) after 15 steps0.571
The transition matrix is T = [[P(A→A), 1−P(B→B)], [1−P(A→A), P(B→B)]]. For a regular chain (not both P(A→A) and P(B→B) at the same extreme), whatever starting P(A) you choose, the curve settles onto the dashed steady-state line πA = (1−P(B→B)) / [(1−P(A→A))+(1−P(B→B))] — an eigenvector of T for eigenvalue 1, exactly as defined above. Try starting P(A) = 0 vs 1 with the same P(A→A), P(B→B): different paths, same destination. Push both sliders to the same edge and that regularity breaks: at P(A→A) = P(B→B) = 1 every state is absorbing, so the chain never moves at all and 'steady state' is just wherever you started; at P(A→A) = P(B→B) = 0 the chain swaps states every single step and never settles — it oscillates between P(A) and 1−P(A) forever.
● Graph theory (networks)
A graph consists of vertices and edges. Weighted graph: edges have assigned weights (distances, costs). Key algorithms:
Kruskal's algorithm — minimum spanning tree (MST): add edges in increasing weight, skip cycles. Prim's algorithm — MST from a vertex; add cheapest edge to tree at each step.
Dijkstra's algorithm — shortest path from a source vertex.
Euler circuit: traverses every edge exactly once; requires all vertices to have even degree.
Hamiltonian path: visits every vertex exactly once.
⊙ Explorer · Kruskal's algorithm, step by stepinteractive
Each edge in sorted order, and whether Kruskal's algorithm added or skipped it
edge
weight
decision
MST weight so far10
edges in tree4 of 4 needed
This is exactly the graph from the Practice Problem below (AB=3, BC=2, CD=4, DE=1, AC=7, BD=5, CE=6, AD=8), sorted by weight: DE, BC, AB, CD, BD, CE, AC, AD. Drag the slider to the right one step at a time — green solid edges were added (they connected two previously-separate pieces of the tree), grey dashed edges were skipped (both endpoints were already connected, so adding one would create a cycle). A spanning tree on 5 vertices always needs exactly 4 edges — once the count reaches 4, every later edge in the list is necessarily skipped, whatever weight it has.
Column-Stochastic vs Row-Stochastic: The IB AI HL syllabus uses column-stochastic transition matrices where each column sums to 1: entry Tij represents the probability of moving from state j to state i. Multiplying Tpn=pn+1 requires a column vector. If you set up rows summing to 1 by mistake, standard matrix multiplications produce wrong states.
Cycle Detection in Kruskal's: In Kruskal's algorithm, check for cycles across the entire connected subgraph, not just whether both vertices have been visited. If two visited vertices already belong to the same component, connecting them creates a cycle and the edge must be rejected.
Eulerian Graph Conditions: An Eulerian trail (open path visiting every edge once) requires exactly two vertices of odd degree. An Eulerian circuit (closed loop returning to the start) requires all vertices to have even degree.
Practice: A network has edges: DE=1, BC=2, AB=3, CD=4, BD=5, CE=6, AC=7, AD=8. Use Kruskal's algorithm to determine the minimum spanning tree and its total weight.
Sort edges: DE(1), BC(2), AB(3), CD(4), BD(5)... Add DE, BC, AB, CD without cycles (4 edges connect 5 vertices). Total weight = 1 + 2 + 3 + 4 = 10.
Practice: A network has weighted adjacency matrix W. Explain how the minimum spanning tree algorithm (Kruskal's) differs from Dijkstra's algorithm.
Kruskal's builds the MST by always adding the globally cheapest edge that does not form a cycle. Dijkstra's finds shortest paths from a single source, greedily updating distance estimates.
📝 Chapter Quiz
§
Practice Set
1Easy
Express z=−1+i3 in polar form r(cosθ+isinθ) and find z6.
r=∣z∣=1+3=2,θ=π−3π=32π. So z=2(cos32π+isin32π). By De Moivre: z6=26(cos4π+isin4π)=64(1+0i)=64.
2Easy
Prove by induction that 5n−1 is divisible by 4 for all positive integers n.
Base:n=1: 51−1=4. ✓ Inductive step: Assume 4∣(5k−1), i.e. 5k=4m+1. Then 5k+1−1=5⋅5k−1=5(4m+1)−1=20m+4=4(5m+1). ✓ By induction, 4∣(5n−1) for all n∈Z+.
Integration by parts: u=ln(x+1),dv=xdx⇒du=x+11dx,v=2x2. =[2x2ln(x+1)]01−∫012(x+1)x2dx. First term: 21ln2−0=2ln2. For the integral, write x+1x2=x−1+x+11, so ∫012(x+1)x2dx=21[2x2−x+ln(x+1)]01=21(21−1+ln2)=2ln2−21. Answer: 2ln2−2ln2−21=2ln2−2ln2+41=41.
Solve the differential equation dxdy=2xy2 with initial condition y(0)=1.
Separate variables: y2dy=2xdx. Integrate: −y1=x2+C. Apply IC: −1=0+C⇒C=−1. So −y1=x2−1⇒y=1−x21,∣x∣<1.
7Medium
Prove that 2 is irrational.
Proof by contradiction: Suppose 2=qp in lowest terms (gcd(p,q)=1). Then 2=q2p2⇒p2=2q2. So p2 is even, meaning p is even. Write p=2k. Then 4k2=2q2⇒q2=2k2, so q is also even. This contradicts gcd(p,q)=1. Therefore 2 is irrational. □
8Medium
Find the equation of the plane passing through A(1,0,2),B(3,1,0),C(0,2,1).
AB=21−2,AC=−12−1. Normal n=AB×AC=i2−1j12k−2−1=(−1+4)(2+2)⋅(−1)(4+1) n=3−45. Using point A(1,0,2): 3(x−1)−4(y)+5(z−2)=0⇒3x−4y+5z=13.
9Hard
Use the binomial theorem to find the coefficient of x4 in the expansion of (2x−x1)8.
General term: (r8)(2x)8−r(−x1)r=(r8)28−r(−1)rx8−r−r=(r8)28−r(−1)rx8−2r. For x4: 8−2r=4⇒r=2. Coefficient: (28)⋅26⋅(−1)2=28×64×1=1792.
10Hard
Find all complex roots of z3=−8 and plot them on the Argand diagram (describe positions).
Write −8=8eiπ. The three cube roots are zk=2ei(π+2πk)/3 for k=0,1,2. z0=2eiπ/3=2(21+i23)=1+i3. z1=2eiπ=−2. z2=2ei5π/3=2(21−i23)=1−i3. On the Argand diagram, the three roots lie on a circle of radius 2, equally spaced at 120° apart, with z1=−2 on the negative real axis.
11Hard
Solve the second-order ODE y′′−5y′+6y=0 with y(0)=1,y′(0)=3.
Characteristic equation: m2−5m+6=0⇒(m−2)(m−3)=0⇒m=2,3. General solution: y=Ae2x+Be3x. y(0)=1⇒A+B=1. y′=2Ae2x+3Be3x,y′(0)=3⇒2A+3B=3. Solving: A=0,B=1. So y=e3x.
12Hard
(IB AI HL) A transition matrix T=(0.70.30.20.8) represents weekly customer movement between two brands. Find the steady-state distribution.
Find eigenvector for eigenvalue λ=1. Solve (T−I)x=0: (−0.30.30.2−0.2)(x1x2)=0⇒−0.3x1+0.2x2=0⇒x2=23x1. Normalise: x1+x2=1⇒x1+23x1=1⇒x1=52,x2=53. Steady-state: Brand A holds 40%, Brand B holds 60% of customers.
✎
Test Generator
Generate a full IB Math HL exam
Creates Paper 1 (no GDC) + Paper 2 (GDC) + Paper 3 (GDC, two extended problem-solving questions) following the IB AA HL format with short-response and extended-response questions.
📝 Chapter Quiz
↓
Downloads & Resources
IB Mathematics HL — Key resources
Paper 3 is the extended investigation paper (60 min). It assesses problem-solving, mathematical communication, and exploration — no specific syllabus topics are assigned to Paper 3.
HL Formula Booklet
More extensive than SL; includes series, complex number forms, vector formulae, and distribution tables.
AA HL Past Papers
Available 2019–present (new syllabus). Focus on proof questions, calculus, complex numbers, and vectors.
AI HL Past Papers
Available 2021–present. Focus on statistics, modelling, matrices, and network problems.
Paper 3 Preparation
Practice with past Paper 3 questions; focus on mathematical communication and justification. Typically 2 questions, each 25–30 marks.
Official · Past Papers
Past Papers & Official Tests
Officially released exams and mark schemes from the examining body.
14 original ClipSAT worksheets with full answer keys, organized by unit — free to download and print.
Loading worksheet library…
Privacy Policy
Last updated July 14, 2026
ClipSAT is a static study site with no user accounts and no server-side database of your personal information. This page explains what little data does exist and how it's handled.
What's stored on your device
Progress, streaks, mistake log, and dark-mode preference are saved in your browser's local storage. They never leave your device and are not visible to us.
If you use "Ask Mr. Mohamed," your questions are sent directly to a third-party AI provider to generate a reply; they are not stored on any ClipSAT server.
Visitor counter
The home page displays a live, site-wide visit count, stored in ClipSAT's own Supabase project — the same optional backend used for cloud sync, described above. Each browser tab counts itself at most once per visit, by calling a small server-side function that only ever increments one shared number; no page content, device identifiers, or account information is attached to that call. As with any request made over the internet, Supabase's own infrastructure sees the request (including your device's IP address) the way any web host does, but ClipSAT's own code never stores or has access to that information — only the running total the function returns.
Advertising & cookies
ClipSAT may show ads served by Google and other third-party vendors. These vendors, including Google, use cookies to serve ads based on a user's prior visits to this or other websites. Google's use of advertising cookies enables it and its partners to serve ads based on your visit to this site and/or other sites on the Internet. You may opt out of personalized advertising by visiting Google Ads Settings. You can also control cookies through your browser settings; disabling cookies may affect some site features.
Google Forms integration
Teachers can optionally connect a Google account to turn a ClipSAT-generated quiz or worksheet into a Google Form, with a shareable link they can post anywhere — Google Classroom, Canvas, Moodle, or any other LMS. This is entirely opt-in — nothing is sent to Google unless a teacher clicks "Connect Google Account" and approves the specific permissions requested. Key points:
ClipSAT requests only the permissions needed to create a Form, read its own responses, and save the question images it renders to a teacher's Google Drive (a permission scoped only to files ClipSAT itself creates — it cannot see or touch anything else already in that Drive). ClipSAT never reads or changes anything else in a connected Google account, and never requests access to a teacher's Google Classroom courses or roster.
Questions and answer choices containing complex math are rendered into an image (so they display correctly in Forms, which cannot show math notation) and uploaded to that teacher's own Google Drive, set to "anyone with the link can view" so the Form can display it. These images are not stored anywhere by ClipSAT — only in the teacher's own Drive.
This connection happens entirely in your browser. ClipSAT has no backend server, so no quiz content, response data, or access token passes through or is stored on any ClipSAT server — each Google API call is made directly from the teacher's browser to Google, using a token that Google itself issues and that ClipSAT never persists beyond the current browser tab.
Any results shown inside ClipSAT (a per-student score summary) are fetched live from Google each time the page is viewed, not stored anywhere by ClipSAT.
Google's own Privacy Policy governs how Google handles data once ClipSAT's requests reach it — see policies.google.com/privacy.
Links to WhatsApp, Telegram, YouTube, Instagram, TikTok, Facebook, and PayPal take you to those companies' own sites, each governed by its own privacy policy.
Contact
Questions about this policy can be sent via WhatsApp or Telegram.
Notes, explorers, and practice questions are provided for personal study. Downloaded packets may be used for your own review; please don't redistribute or resell them.
No guarantee of results
ClipSAT is a study aid, not a guarantee of any particular exam score. Always check official curriculum documents (e.g. Cambridge, IB, College Board, Qiyas) for the authoritative syllabus.
AI tutor
"Ask Mr. Mohamed" is an AI assistant, not Mr. Mohamed Abdallah himself. Double-check important steps and final answers yourself.
ClipSAT's AI tutor and question generator work for free once you're signed in (☁️ Sign in, top of the page). Paste your own Groq key here instead if you'd rather not sign in, or want a private, higher-limit key — leave it blank and save to go back to the shared sign-in-based one.