AP® CALCULUS BC ONLY · UNIT 7

Approximating Solutions Using Euler’s Method

Follow the current slope, take a step, then repeat. Build accurate Euler tables, compare numerical paths and explain what your approximation means.

Topic 7.5 study notes8 worked examples · 8 practice questionsUpdated 12 November 2026
Calculus notebook with limit and integral symbols, a compass and pencil on an olive and cream study desk

1. The Euler update and AP BC scope

Euler’s method (pronounced “OY-ler”) approximates a solution to an initial value problem by repeatedly following a short straight segment with the slope prescribed by the differential equation at the current point. You do not need a closed-form solution to carry out the method.

AP Calculus BC only · Topic 7.5. The course framework identifies learning objective FUN-7.C, essential knowledge FUN-7.C.4 and suggested skill 1.E: use a numerical procedure to estimate a solution or a point on a solution curve. Euler’s method is not a required AP Calculus AB topic.
dydx=f(x,y)
y(x0)=y0

The initial condition supplies the exact starting point. After the first step, the computed heights are generally estimates. The method still evaluates the given rate at those computed points.

The two updates

xn+1=xn+h
yn+1=yn+hf(xn,yn)

Here h is the signed horizontal step and f(xn,yn) is the slope at the current numerical point. In words: new height equals old height plus slope times horizontal change. The equality defines the numerical update; the final numerical height is an approximation to the true solution value.

y(xn)≈yn

2. Choose the step and organize the process

A reliable Euler workflow
StageWhat to do
StartRead the initial point (x0,y0) and the requested target input b.
Choose the stepFor N equal steps, use h=b−x0N. A target to the left gives a negative step.
Find the slopeCalculate mn=f(xn,yn) using both coordinates from the same current row.
UpdateCompute xn+1=xn+h and yn+1=yn+hmn. Do not use a partially updated point to find the old slope.
Repeat and stopThe new point becomes the current point. Stop when xN=b; N steps give N+1 point rows.
ReportWrite y(b)≈yN, include units when given, and round only the final result as directed.

How to find the step size

h=b−x0N

Four equal steps from 0 to 2 give h=2−04=0.5. If a problem directly specifies h=0.1, use that value. Five equal steps from 1 to 3 give h=3−15=0.4.

When the step size is fixed, b−x0h must be a nonnegative integer to reach the target with that many equal steps. Do not silently take a shorter last step when a problem requires equal steps.

Connection to local linearization

L(x)=yn+f(xn,yn)(x−xn)

Evaluating this line at xn+h produces the next estimate. At the initial exact point it is tangent to the true initial-value solution. At a later computed point it is a direction supplied by the equation; the computed point need not lie on the true solution you started with.

3. Eight colorful worked examples

Example 1 · Four steps from the initial point

Use h=0.5 to approximate y(2) for dydx=x+y, with y(0)=1.

1 · Set up the march

N=2−00.5=4

Start at (0,1). Each step evaluates f(x,y)=x+y at the current point.

2 · Take the first step

m0=0+1=1
x1=0+0.5=0.5
y1=1+0.5(1)=1.5

The next point is (0.5,1.5).

3 · Recompute before the second step

m1=0.5+1.5=2
x2=0.5+0.5=1
y2=1.5+0.5(2)=2.5

4 · Continue from the new point

m2=1+2.5=3.5
x3=1+0.5=1.5
y3=2.5+0.5(3.5)=4.25

5 · Finish at the target

m3=1.5+4.25=5.75
x4=1.5+0.5=2
y4=4.25+0.5(5.75)=7.125
y(2)≈7.125

Four updates have reached the requested input. This is an Euler estimate, not an exact solution value.

Example 2 · Read the same work in a table

Organize Example 1 without changing the differential equation, starting point or step size.

1 · Keep one point on each row

Original four-step calculation, organized by current point
Step nxnynCurrent slope
0011
10.51.52
212.53.5
31.54.255.75
427.125Stop

2 · Read the final row correctly

The last row gives (x4,y4)=(2,7.125). The slope column is marked “Stop” because no fifth update is requested. The arithmetic for every update appears in Example 1.

y(2)≈7.125

A table is a useful organizing choice. Clear labeled calculations also communicate the method.

Example 3 · Three short steps for exponential growth

Use h=0.1 to approximate y(0.3) for dydx=y, y(0)=1.

1 · Count and identify the slope

N=0.3−00.1=3

At each point the slope is the current height. The inputs are 0, 0.1, 0.2 and 0.3.

2 · Use each new height on the next step

y1=1+0.1(1)=1.1
y2=1.1+0.1(1.1)=1.21
y3=1.21+0.1(1.21)=1.331

3 · Label the estimate

y(0.3)≈1.331

For the supplied exact solution y=ex, the true value is about 1.350. The Euler value is lower in this example.

Example 4 · One step is one local linear estimate

Given y(1)=3 and dydx=2x, use one step of size 0.5 to estimate y(1.5).

1 · Evaluate the starting slope

m0=2(1)=2

2 · Multiply by the horizontal step

x1=1+0.5=1.5
y1=3+0.5(2)=4

3 · Compare with a supplied exact solution

y(1.5)≈4

The function y=x2+2 satisfies the initial value problem and has y(1.5)=4.25. Its positive second derivative explains why the starting tangent line lies below this convex solution.

Example 5 · Find the step size from the number of steps

An initial value problem starts at x0=1 and must reach x=3 in five equal steps. Find the step size and input sequence.

1 · Divide the whole horizontal change

h=3−15=25=0.4

2 · List six points for five moves

1, 1.4, 1.8, 2.2, 2.6, 3

The initial point is not a completed step. Five updates produce six point rows.

Example 6 · Move backward with a negative step

For dydx=x+y, y(1)=2, estimate y(0.5) using two equal steps.

1 · Keep the direction in the step size

h=0.5−12=−0.25

2 · Follow the same update rule

m0=1+2=3
y1=2+(−0.25)(3)=1.25
m1=0.75+1.25=2
y2=1.25+(−0.25)(2)
y2=0.75

3 · Check the target

Two backward steps
Step nxnynCurrent slope
0123
10.751.252
20.50.75Stop
y(0.5)≈0.75

A positive slope with a negative horizontal step gives a negative height change. The formula itself is unchanged.

Example 7 · Compare two step sizes at the same target

Repeat Example 1 with h=0.25. Compare the eight-step estimate of y(2) with the four-step estimate.

1 · Apply the same recurrence eight times

yn+1=yn+0.25(xn+yn)

The successive heights, retaining all digits, are 1, 1.25, 1.625, 2.15625, 2.8828125, 3.853515625, 5.12939453125, 6.7867431640625 and 8.920928955078125.

2 · Compare with a verified reference

The supplied exact solution is y=2ex−x−1. Its derivative equals x+y, and it passes through (0,1).

y(2)=2e2−3≈11.778
Step-size comparison at the same target
Step sizeStepsEuler estimateAbsolute error
0.547.1254.653
0.2588.9212.857

3 · State what this comparison establishes

y(2)≈8.921using eight steps

Halving the step improved the result in this example. The estimates are still below the true value. A step-size comparison is evidence for this problem, not a guarantee that every smaller step strictly improves every numerical answer.

Example 8 · Recognize an overestimate in a checked case

Use h=0.5 to estimate y(1) for dydx=2−y, y(0)=0.

1 · Take two forward steps

y1=0+0.5(2−0)=1
y2=1+0.5(2−1)=1.5

2 · Check the error direction

The supplied exact solution is y=2−2e−x. It is concave down, with y(1)=2−2e−1≈1.264. Direct comparison gives an Euler overestimate.

y(1)≈1.5

The positive rates get smaller as the numerical heights rise. Reusing the initial slope for both steps would give a different, incorrect Euler calculation.

4. Six verified numerical paths and reference curves

Blue solid lines join Euler points; filled blue circles are the computed points, not exact solution values. Orange dashed curves are supplied exact solutions used to verify the comparisons. The refinement panel also uses a green dash-dot path with square markers. Each panel uses independent horizontal and vertical scales for readability; compare the labeled coordinates rather than visual angles.

Four Euler segments and the exact curve

image/svg+xml Matplotlib v3.11.2, https://matplotlib.org/ 0.0 0.5 1.0 1.5 2.0 x 1 4 7 10 12 y Exact Euler
For dydx=x+y, y(0)=1, the four-step path reaches (2,7.125). The exact curve y=2ex−x−1 reaches the larger value 2e2−3.

Smaller steps in the same problem

image/svg+xml Matplotlib v3.11.2, https://matplotlib.org/ 0.0 0.5 1.0 1.5 2.0 x 1 4 7 10 12 y Exact Euler Finer Euler
The blue path uses h=0.5; the green dash-dot path uses h=0.25. Both start at (0,1) and stop at x=2. The eight-step endpoint is (2,8.920928955078125), closer to the orange exact curve.

A single starting tangent segment

image/svg+xml Matplotlib v3.11.2, https://matplotlib.org/ 1.00 1.25 1.50 x 3.00 3.50 4.00 4.25 y Exact Euler
For dydx=2x, y(1)=3, the Euler endpoint is (1.5,4). The orange curve y=x2+2 is concave up and reaches (1.5,4.25).

Two steps toward smaller inputs

image/svg+xml Matplotlib v3.11.2, https://matplotlib.org/ 0.50 0.75 1.00 x 0.75 1.00 1.50 2.00 y Exact Euler
For dydx=x+y, y(1)=2, arrows follow the computed path from right to left: (1,2), (0.75,1.25), (0.5,0.75). The exact reference is y=4ex−1−x−1.

Three updates for a decreasing solution

image/svg+xml Matplotlib v3.11.2, https://matplotlib.org/ 0.0 0.2 0.4 0.6 x 1.2 1.4 1.6 1.8 2.0 y Exact Euler
For dydx=x−y, y(0)=2, the step size 0.2 produces y(0.6)≈1.136. The exact reference is y=x−1+3e−x.

Finish all four product-rate steps

image/svg+xml Matplotlib v3.11.2, https://matplotlib.org/ 0.0 0.5 1.0 1.5 2.0 x 1 3 5 7 y Exact Euler
For dydx=xy, y(0)=1, the four-step Euler estimate is y(2)≈3.28125. The orange exact curve is y=ex22; the second Euler height 1.25 is not the final answer.

The plotted intervals are the requested approximation intervals. The endpoints of the numerical paths are the start and target, not domain endpoints of the exact solutions; all six supplied exact solutions are defined for every real input. Abstract coordinates have no specified physical units. A later Euler segment follows the rate at a computed point and is generally not tangent to the original exact curve.

5. Accuracy, concavity and domain checks

Approximation, rounding and error propagation

Euler’s method replaces curved motion with straight steps. Numerical and rounding errors can propagate through later updates; errors need not grow strictly at every step and may cancel. Keep full intermediate precision, store values carefully on a calculator, and follow the problem’s final rounding instruction. Exact fractions are also useful for checking arithmetic.

For sufficiently regular problems on a fixed interval, taking sufficiently small steps improves convergence to the true solution. A large step can be inaccurate or behave badly. This topic focuses on carrying out the requested Euler process; a formal error bound or proof of convergence is not needed for the worked calculations here.

Use concavity carefully

Starting at an exact point, a tangent-line estimate lies below a solution that is concave up across the step, and above a solution that is concave down across the step. For a multistep Euler calculation, later starting points are approximate. Do not automatically call the entire path an overestimate or underestimate from a sketch or one initial concavity value; justify the error direction for the specified equation and interval, or compare with a valid reference solution.

Examples 4 and 8 verify those tangent and error comparisons directly. Example 7 checks the actual final errors at a shared target. A plotted gap is an illustration; the numerical calculation supplies the quantitative comparison.

Check the equation’s domain

Every slope evaluation must use a point where the right side is defined. Do not evaluate a zero denominator, step across an excluded input as though the equation were defined there, or infer a true solution’s global existence from a finite list of numerical points.

Useful special forms

If dydx=ky, a fixed step gives

yn+1=(1+kh)yn

If dydx=kx, the height change on the current step is hkxn. A constant derivative gives an exact straight-line solution, so its Euler values can be exact in exact arithmetic. “Numerical method” does not mean every possible result must have nonzero error.

6. Eight practice questions with complete solutions

1. Update a decreasing numerical solution

Use h=0.2 to approximate y(0.6) for dydx=x−y, y(0)=2.

Show complete solution

1 · Recompute every slope

y1=2+0.2(−2)=1.6
y2=1.6+0.2(−1.4)=1.32
y3=1.32+0.2(−0.92)
y3=1.136

2 · Match the final row to the target

All three steps for the original first practice question
Step nxnynCurrent slope
002−2
10.21.6−1.4
20.41.32−0.92
30.61.136Stop
y(0.6)≈1.136

Check all three current slopes before accepting the final estimate. The final input is 0.6, after exactly three updates.

2. Use one growth step

Use one step with h=0.5 to approximate y(2.5) for dydx=2y, y(2)=3.

Show complete solution

1 · Evaluate the given rate

m0=2(3)=6

2 · Take the requested step

x1=2+0.5=2.5
y1=3+0.5(6)=6
y(2.5)≈6

3. Finish all four product-rate steps

Use four equal steps to approximate y(2) for dydx=xy, y(0)=1.

Show complete solution

1 · Find the step and keep the zero first slope

h=2−04=0.5
y1=1+0.5(0)=1
y2=1+0.5(0.5)=1.25

2 · Complete the remaining two steps

y3=1.25+0.5(1.25)=1.875
y4=1.875+0.5(2.8125)
y4=3.28125

3 · Report the fourth height

All four steps for the original third practice question
Step nxnynCurrent slope
0010
10.510.5
211.251.25
31.51.8752.8125
423.28125Stop
y(2)≈3.28125

The value 1.25 occurs at x=1, only halfway to the target. It is not the answer at x=2.

4. Recognize an exact straight-line case

Use h=0.4 to approximate y(1.2) for dydx=3, y(0)=1.

Show complete solution

1 · Add the same rise three times

y1=1+0.4(3)=2.2
y2=2.2+0.4(3)=3.4
y3=3.4+0.4(3)=4.6

2 · Check against the linear solution

The exact solution is y=1+3x. Its value at 1.2 is exactly 4.6, so the Euler estimate is exact in this case.

y(1.2)=4.6

5. Use a specified number of steps

For dydx=x−y, y(1)=2, choose five equal steps from 1 to 3. Find the first new point.

Show complete solution

1 · Compute the step and current slope

h=3−15=0.4
m0=1−2=−1

2 · Update both coordinates

x1=1+0.4=1.4
y1=2+0.4(−1)=1.6

The first numerical point is (1.4,1.6). Four further steps would be needed to reach 3.

6. Take two negative steps

For dydx=2x, y(1)=3, approximate y(0) with two equal steps.

Show complete solution

1 · Move left by half a unit per step

h=0−12=−0.5

2 · Recompute the rate at the next input

y1=3+(−0.5)(2)=2
y2=2+(−0.5)(1)=1.5
y(0)≈1.5

The intermediate inputs are 1, 0.5 and 0. The second slope is 2(0.5)=1.

7. Diagnose a partially updated slope

For dydx=x+y, y(0)=1, a student uses the point (0.5,1) to find the first slope for h=0.5. Correct the step.

Show complete solution

1 · Use the current pair before changing either coordinate

m0=f(0,1)=1

The mixed point (0.5,1) has a new input and an old height. It is not the starting pair.

2 · Apply the correct old slope

x1=0.5
y1=1+0.5(1)=1.5
y(0.5)≈1.5

8. Keep rates and units in a cooling model

A model uses dTdt=−0.1(T−20), with T(0)=80. Time is in minutes and temperature in degrees Celsius; the coefficient has units min−1. Use two one-minute steps to estimate the temperature after two minutes.

Show complete solution

1 · Find the first rate and temperature

The initial rate is −6∘Cmin.

T1=80+1(−6)=74

2 · Use the updated temperature

m1=−0.1(74−20)=−5.4
T2=74+1(−5.4)=68.6
T(2)≈68.6∘C

The numerical temperatures decrease while remaining above the ambient temperature in these two steps.

7. Common mistakes and clear written work

  • Omitting the step size. A slope is a rate, not the height change. Multiply by h.
  • Using a mixed point. Compute f(xn,yn) before updating either coordinate.
  • Reusing an obsolete height or slope. Once a step is complete, use the new pair for the next rate evaluation.
  • Confusing the step index with the input. The label n counts updates; xn records the actual input.
  • Stopping one row early. Include the initial row and verify the target input before reporting the final height.
  • Rounding repeatedly. Retain extra digits or exact fractions until the end; check multiplication and addition separately.
  • Losing the direction. A target to the left requires a negative step.
  • Using an undefined rate. Check original denominator restrictions and the equation’s domain before evaluating slopes.
  • Claiming an exact answer without justification. Use an approximation sign unless an exact case has been verified.
  • Skipping the method. Show the requested step size, rate evaluations and updates rather than only an unexplained final decimal.

A clear written response

Identify the initial condition, write the update with the requested step size, show each evaluated slope or a labeled table, verify the final input, then state the estimated value with units when appropriate. Follow the precision requested in the question. These habits make the mathematics checkable; the points on any particular AP question depend on that question’s scoring guidelines.

8. Frequently asked questions

Is Euler’s method required for AP Calculus AB?

No. Topic 7.5 is marked BC only in the AP Calculus course framework. The earlier slope-field topics are shared by AB and BC.

Do I need to solve the differential equation first?

No. The given rate and initial point are enough for the numerical updates. Exact solutions on this page are supplied to check selected examples and graphs.

Which point supplies the next slope?

Use the current numerical pair (xn,yn). After both updates are finished, the new pair supplies the slope for the following step.

How many rows should the table have?

For N steps, include N+1 point rows: the initial point and one new point per update. An extra slope at the final target is unnecessary unless another step is requested.

Can the step size be negative?

Yes. Use h=b−x0N. A smaller target input makes the step negative, and the same recurrence applies.

Do smaller steps always give a strictly better answer?

Not as an unconditional claim for every individual step size and problem. With suitable regularity and sufficiently small steps, refinement improves convergence. Compare at the same target, retain precision, and check the actual equation.

When should I round and how should I report the answer?

Keep intermediate digits or exact fractions. Round the final value as directed and write y(b)≈yN, with units if the variables have physical meaning. Some linear cases can be exact and should be identified with a valid check.

9. Continue through Unit 7

Euler’s method turns the local directions in a slope field into a numerical path. Continue to separation of variables to learn when an equation can also be solved symbolically.

Use the numerical update when an Euler approximation is requested, even if an exact solution is available. The exact solution and the requested numerical estimate answer different instructions.