The inverse function theorem explains how the derivative of a function is related to the derivative of its inverse. It matters because inverse functions appear throughout calculus, including logarithms, inverse trigonometric functions, and solving equations by reversing a process. The central idea is that if a function has a nonzero slope at a point and is locally invertible there, then its inverse is also differentiable at the matching point.
On a graph, a function and its inverse are mirror images across the line y = x.
Understanding Calculus: The Inverse Function Theorem
The result comes from composing a function with its inverse. Start with an input near a chosen value. The function sends it to an output, then the inverse sends that output straight back to the original input.
This combined process acts like the identity function, whose rate of change is one. The chain rule says that the rate of the first step times the rate of the return step must equal one.
Therefore, the inverse rate must be one divided by the original rate. This argument explains the rule rather than treating it as a fact to memorize.
The word local is important. A function does not need to reverse perfectly over its whole graph. It only needs to reverse on a small interval around the point being studied.
For example, the squaring function cannot have one inverse on all real numbers because positive and negative inputs can give the same output. Near a positive input, however, it behaves in only one direction and can be reversed by the positive square root. Near zero, its slope is zero.
The square root then has an infinitely steep graph at the matching output. This is why a zero derivative prevents the usual inverse derivative rule from working.
This idea saves work when an inverse has no convenient formula. Suppose a function is defined by a complicated expression or by measured data. To find how fast the input changes when the output changes, first locate the input that produces the output of interest.
Next find the original function's derivative at that input. Finally take its reciprocal. This is closely related to implicit differentiation.
In both methods, students find a rate from a relationship without fully solving that relationship for one variable. The method appears when studying inverse sine, inverse tangent, logarithms, roots, and many functions defined only by equations.
Reciprocal slopes have a clear physical meaning. If a sensor output changes very quickly when its input changes, then reversing the sensor reading gives a small input change for each output change. If the original slope is close to zero, the reversed process becomes highly sensitive.
Small errors in output can create large errors in the recovered input. This matters in calibration, computer graphics, engineering measurements, and scientific models. When practicing, keep track of which point belongs to the original function and which belongs to the inverse.
The input and output switch places. Students often take the derivative at the given output by mistake, rather than first finding the original input that produced it.
Key Facts
- If f is differentiable near a, f'(a) ≠ 0, and f has an inverse near a, then (f^-1)'(f(a)) = 1 / f'(a).
- If b = f(a), then (f^-1)'(b) = 1 / f'(a).
- The slope of the inverse at (f(a), a) is the reciprocal of the slope of f at (a, f(a)).
- A nonzero derivative f'(a) means the tangent line is not horizontal, so the reflected inverse tangent is not vertical.
- The graphs of y = f(x) and y = f^-1(x) are reflections across y = x.
- To find (f^-1)'(b), solve f(a) = b first, then compute 1 / f'(a).
Vocabulary
- Inverse function
- An inverse function reverses the input and output of a function, so f^-1(f(x)) = x for allowed values of x.
- Local inverse
- A local inverse is an inverse that exists only on a small interval near a chosen point.
- Derivative
- The derivative gives the instantaneous rate of change or slope of a function at a point.
- Reciprocal slope
- A reciprocal slope is formed by switching numerator and denominator, such as changing m into 1/m.
- Reflection across y = x
- Reflection across y = x swaps every point (x, y) on a graph with the point (y, x).
Common Mistakes to Avoid
- Using (f^-1)'(a) = 1 / f'(a) directly is wrong because the input to the inverse derivative must be an output value of f. If b = f(a), then the correct formula is (f^-1)'(b) = 1 / f'(a).
- Forgetting to solve f(a) = b first is wrong because the derivative f'(a) must be evaluated at the original x-value, not at the inverse input b.
- Applying the theorem when f'(a) = 0 is wrong because the reciprocal 1 / f'(a) is undefined and the inverse may fail to be differentiable there.
- Assuming a global inverse always exists is wrong because a function may only be one-to-one on a restricted interval. The theorem only needs an inverse near the point.
Practice Questions
- 1 Let f(x) = x^3 + x. Find (f^-1)'(2).
- 2 Let f(x) = e^x + 2x. Find (f^-1)'(1).
- 3 A differentiable function has f(3) = 7 and f'(3) = -4. Explain what point lies on the graph of f^-1 and what the slope of f^-1 is at that point.