Calculus
18 lessons · 10 HL only · 136 past-paper questions in the AI HL bank
● SL AND HL
8 lessonsA limit is the value a function's output heads for as its input gets close to some number. Limits turn the gradient of a chord into the gradient of a curve at a single point. That gradient, written f'(x) or dy/dx, is the derivative: it tells you how steep a curve is and how fast a quantity is changing, such as users per week or profit per extra item.
A function is increasing where its graph rises from left to right, and decreasing where it falls. The sign of the derivative tells you which: f'(x) > 0 means rising, f'(x) < 0 means falling, and f'(x) = 0 means a flat tangent at a stationary point. In a model, this tells you when a quantity such as a population, a profit or a temperature is going up or down.
The power rule turns axⁿ into anxⁿ⁻¹: multiply by the power, then lower the power by one. It works for every integer power, including negative ones such as 5/x² = 5x⁻². Differentiate a sum one term at a time, and you can find the rate of change of any polynomial model, such as a cost or a temperature, in a few lines.
A tangent is the straight line that touches a curve at a point and has the curve's gradient there. The normal goes through the same point at right angles to the tangent. For either line you need the point, from f(a), and the gradient, from f'(a), which you find by differentiating or with your GDC.
Integration runs differentiation backwards: given a rate of change, you rebuild the function, then use one known point to pin down its constant. A definite integral, found on your GDC, turns a rate into a total change. For a curve above the x-axis, it is also the area under the curve.
At a local maximum or local minimum inside the domain of a smooth curve, the gradient is zero, so you find these points by solving f'(x) = 0. On Paper 2 your GDC can draw y = f'(x) from f(x) and solve the equation for you. A local maximum need not be the greatest value on the domain: check the end points too.
Optimisation means finding the best value of something in a real situation: the greatest profit, the least cost, the largest volume. Write the quantity as a function of one variable, solve where its derivative is zero, then check that point against the ends of the allowed values. Always finish by answering the question in context.
The trapezoidal rule estimates the area under a curve. It splits the area into strips of equal width and treats each strip as a trapezoid with a straight top. It works from a table of measurements or from a function, and when the curve bends the same way throughout, its shape tells you whether the estimate is too big or too small.
◆ HL ONLY
10 lessonsSix standard derivatives, for xⁿ, sin x, cos x, tan x, eˣ and ln x, plus three rules let you differentiate most models in this course. The chain rule handles a function inside another, and the product rule and quotient rule handle products and fractions. The chain rule also links related rates, and every rule helps you find maximum and minimum points.
The second derivative, f''(x), is the derivative of the derivative: it tells you how fast the gradient is changing. At a stationary point, if it is not 0, its sign tells a maximum from a minimum. Elsewhere its sign tells you which way the graph bends, and where the bend changes there is a point of inflexion, often where a quantity grows or falls fastest.
Read any derivative backwards and you have an integral. You'll integrate xⁿ for rational n (including 1/x), sin x, cos x, 1/cos²x and eˣ, and find definite integrals exactly. Then you'll undo the chain rule, by inspection or substitution, to find totals from rates.
A definite integral counts area below an axis as negative, so an area adds the sizes of the parts. Turn a region once about the x- or y-axis and it sweeps out a solid, whose volume of revolution adds up thin discs.
Displacement s, velocity v and acceleration a are linked by calculus: differentiate to go from s to v to a, and integrate to go back. Speed is the size of v. Distance travelled counts every metre, either way, so it integrates |v|.
A differential equation says how fast something changes, such as "the drug leaves the blood at a rate proportional to the amount left". You'll turn sentences like that into equations, then solve them by separation of variables. The answer is a model you can use to predict and interpret.
A slope field draws a short line at each grid point with the gradient that a differential equation gives there. Solution curves run along these lines, so you can sketch a solution and see its long-term behaviour without solving the equation. In a model, the field shows where a quantity rises, falls, peaks or levels off.
Euler's method estimates the solution of a differential equation that you can't solve exactly. From a known point, you follow the gradient for a short step, then repeat. A spreadsheet or GDC does the repeating, for one equation or for a pair of coupled differential equations such as a predator–prey model.
A phase portrait shows where a pair of linear coupled differential equations takes every starting point. The eigenvalues of the system's matrix set the picture: paths in or out, spirals, ellipses or a saddle point. You use it to predict long-term behaviour, such as whether a population settles or dies out.
A second-order differential equation gives an acceleration, d²x/dt², such as for a spring or a bouncing diving board. Put y = dx/dt and it becomes two coupled differential equations. Then you can use Euler's method, find an exact solution from eigenvalues, or read the motion from a phase portrait.