A graph of a derivative is not a graph of the original function. Its height represents a slope, and its signed area represents a change in the original function. Keeping those meanings straight lets you recover a surprising amount of information without knowing a formula for f.
Translate each graph into words
On the graph of , height means output and slope means On the graph of height means the slope of , while slope means Therefore increases where the graph of is above the axis and decreases where it is below.
Concavity comes from whether rises or falls. Where the graph of has positive slope, and is concave up. Where it has negative slope, is concave down. An increasing derivative can have an isolated zero slope without changing the overall concavity. The derivative can be negative and rising simultaneously: then decreases, but its slopes become less negative.
Example: read a derivative sign pattern
Suppose Its graph crosses the axis at and 1. It is positive outside those zeros and negative between them. Thus increases before , decreases between and 1, and increases after 1.
The positive-to-negative change at gives a local maximum. The negative-to-positive change at 1 gives it a local minimum. Meanwhile, the derivative graph decreases for and increases for so changes from concave down to concave up at zero.
One possible original function is Adding any constant gives the same derivative. Consequently, a derivative graph alone cannot tell you the absolute height of . It determines changes and shapes, but one known function value is needed to fix the vertical position.
This is f′, not f. Above the axis means f increases; the rising right half means f is concave up.
Describe this figure
Function values for the solid curve:
| x | f(x) |
|---|---|
| -2 | 3 |
| -1 | 0 |
| 0 | -1 |
| 1 | 0 |
| 2 | 3 |
The axes answer different questions
A zero of is a horizontal-tangent candidate for ; it need not be a zero of . A local maximum of is a candidate for an inflection point of , because changes from increasing to decreasing there. It need not be a maximum of .
For the derivative touches the axis at zero without changing sign. The original graph has a horizontal tangent but keeps increasing. Crossing versus merely touching the axis matters when classifying extrema from a derivative graph.
Example: recover function values from area
Suppose The graph of encloses area 3 above the axis from 0 to 2 and area 4 below it from 2 to 5. The Fundamental Theorem of Calculus gives
The unsigned total area is 7, but the net change is . Below-axis area contributes negatively because it represents decreasing function values. If you were asked for total variation along the graph's output, you would use instead.
To find the largest value of from such a graph, compare accumulated signed areas at endpoints and points where changes from positive to negative. Do not choose the highest point on the derivative graph: that identifies the greatest slope, not necessarily the greatest output.
Try it yourself
At a point, a graph of is below the axis and sloping upward. Is increasing or decreasing? Is it concave up or down? Then suppose crosses from negative to positive at that point.
Answer: Below the axis means so decreases. Upward slope means so is concave up. If crosses from negative to positive, changes from decreasing to increasing and has a local minimum at the crossing, provided is continuous there.
What to remember
Derivative height controls increasing or decreasing behavior; derivative slope controls concavity; derivative signed area controls net change. Always name which function's graph you are reading.