Flow Lab

Syllabus / Module 3: First-Order ODEs

Lesson 3.3

Separable Equations

Separation of variables, implicit solutions, exponential growth/decay, and lost solutions.

Written byClaude Sonnet

🌊
Why this matters for simulation: Viscous decay of a vortex and drag-limited speed are separable problems.

By the end of this lesson you can…

  • Separate the variables in y' = f(t) g(y) and justify the step with the chain rule instead of treating dy/dt as a fraction.
  • Leave a solution implicit or solve it for y, and state the interval on which each form is valid.
  • Solve exponential growth and decay, and convert a rate constant into a half-life or doubling time.
  • Find the constant solutions that dividing by g(y) discards, and decide whether the general formula can recover them.
  • Derive the blow-up time of y' = y^2, and say when a separable equation reaches infinity in finite time.
  • Check an RK4 run against an exact solution, including at the point where the exact solution stops existing.
On this page
·
  1. Separating the variables
  2. Implicit versus explicit
  3. Exponential growth, decay and the half-life
  4. The solutions you lose by dividing
  5. Domain of validity and finite-time blow-up
  6. Analytic against RK4
  7. Exercise: symbolic against numerical
  8. Where this goes next
  9. Key Takeaways

So far you could only draw an ODE (3.2 Direction Fields & Euler's Method Revisited) or march along it with a solver. This lesson gives you the first technique that returns a formula. For an equation of the right shape the curves in the direction field can be written down exactly, and that changes how you use the solvers you built: they are no longer the only source of answers, they are things you can test. A closed form is an answer key.

It also settles an open account. In the last lesson Euler on y′=y2y' = y^2 from (0,0.5)(0, 0.5) reported 1.831.83 at t=2t = 2, where we claimed the true solution had already blown up. We asserted that t=2t = 2; here you derive it by hand, in four lines.

▶ Video Separating variables, the equilibria that division loses, and a solution that ends in finite timecaptions only
Transcript
  1. Last lesson we could draw an equation, or march along it. Now we get a formula, and a way to catch a solver out.
  2. A separable equation has a right side that factors: a function of t times a function of y. Decay is separable. So is minus t over y. But y prime equals t minus y is not. It will not factor.
  3. The recipe treats dy by dt as a fraction, but it is a limit. Why does it work? Divide by g, and let H antidifferentiate one over g. By the chain rule, the t derivative of H of y is y prime over g, which is f of t. So H of y and the integral of f differ by a constant. Note we divided by g.
  4. Try y prime equals minus t over y, with y of zero equal to two. Separating gives y dy equals minus t dt, so y squared plus t squared equals four. That relation is a whole circle.
  5. But a solution must be a function of t. The start y of zero equals two picks the upper half, root of four minus t squared. At t equals plus or minus two the circle turns vertical, the slope is undefined, and the solution ends.
  6. Now y prime equals y squared minus one. Dividing by y squared minus one is illegal at one and minus one, and both are constant solutions. Between them the right side is negative, so solutions fall. Above one and below minus one it is positive, so they rise.
  7. The rest gives y equals one plus K e to the two t, over one minus K e to the two t. K zero recovers y equals one. But y equals minus one needs K infinite. Stop at the formula and you would say that start has no solution.
  8. From y of zero equals two, K is a third, and the denominator vanishes at t equals a half log three, 0.549. The curve is already past six and off the top of this window at 0.38; the wall is where it becomes infinite. From zero, it slides toward minus one, and never reaches it.
  9. Now y prime equals y squared from one half. Separating gives y equals one over two minus t. What runs out is time: the time to climb from one half to infinity is the integral of dy over y squared, which is two. Infinity arrives at t equals two.
  10. RK4 with h a quarter tracks it: three point nine seven two against exactly four at t equals one point seven five. Then the pole is inside its next step: thirty two point eight at t equals two, and four times ten to the eleven after that.
  11. Euler with h a half sails straight past the pole, still reporting small numbers: one point eight two six at t equals two. A finite number is not a solution.
  12. Back to minus t over y, now with y of zero equals one. The exact solution is the upper half of the unit circle, ending at t equals one. Since y squared plus t squared is constant along solutions, its drift measures a solver's error.
  13. Run RK4 with h of a tenth. Up to t equals one it hugs the circle, and the invariant reads one. There the exact solution reaches y equals zero. RK4 lands at 0.049, where the slope is minus twenty point five, and the next step is thrown across the axis. The invariant hops: one point five five, one point seven seven, two point four, then eight point one nine, and stays.
  14. Look at y alone: 2.37, 2.22, 2.05, 1.83. Smooth, finite, no warning. The invariant reads eight point one nine, not one. The values look innocent. The invariant catches it.
  15. Two separable laws inside fluid solvers. A shear wave decays at rate nu k squared, so doubling the wavenumber quadruples the rate: small scales die first. With quadratic drag, speed creeps toward the terminal speed, a lost constant solution: 37.7 at five seconds, 48.1 at ten.
  16. So: separate, justify with the chain rule, and keep the roots of g, the equilibria that dividing loses. The interval of validity belongs to the solution, and a closed form lets you test a solver. Next: linear equations.

Separating the variables

An equation is separable if the right side factors into a function of tt times a function of yy:

y′=f(t) g(y).y' = f(t)\,g(y) .

Exponential decay y′=−kyy' = -ky is separable (f=−kf = -k, g=yg = y), and so is Newton cooling written for the temperature gap. The equation y′=t−yy' = t - y from last lesson is not: no factorization f(t) g(y)f(t)\,g(y) equals t−yt - y. That one needs a different technique (3.4 Linear First-Order Equations & Integrating Factors).

The textbook recipe is to treat dy/dtdy/dt as a fraction: move all the yy‘s to one side, dy/g(y)=f(t) dtdy/g(y) = f(t)\,dt, and integrate both sides. Leibniz notation is built so that this works, but dy/dtdy/dt is a limit, not a quotient, so the recipe needs a reason. Here it is.

Suppose y(t)y(t) solves the equation on an interval where g(y(t))≠0g(y(t)) \neq 0. Divide by gg:

y′(t)g(y(t))=f(t).\frac{y'(t)}{g(y(t))} = f(t) .

Let H(y)H(y) be any antiderivative of 1/g(y)1/g(y). By the chain rule,

ddt H(y(t))=H′(y(t)) y′(t)=y′(t)g(y(t))=f(t).\frac{d}{dt}\,H\big(y(t)\big) = H'\big(y(t)\big)\,y'(t) = \frac{y'(t)}{g(y(t))} = f(t).

So H(y(t))H(y(t)) and F(t)=∫f dtF(t) = \int f\,dt have the same derivative on an interval, hence differ by a constant:

H(y(t))=F(t)+C.H\big(y(t)\big) = F(t) + C .

That is the recipe’s output, reached without ever splitting dy/dtdy/dt. The “fraction” step is the substitution rule ∫1g(y(t)) y′(t) dt=∫dyg(y)\int \frac{1}{g(y(t))}\,y'(t)\,dt = \int \frac{dy}{g(y)} in disguise. The argument also shows exactly where the danger is: we divided by g(y(t))g(y(t)).

Implicit versus explicit

Take y′=−t/yy' = -t/y with y(0)=2y(0) = 2. Here g(y)=1/yg(y) = 1/y, so y dy=−t dty\,dy = -t\,dt and

y22=−t22+C⟹y2+t2=4,\frac{y^2}{2} = -\frac{t^2}{2} + C \qquad\Longrightarrow\qquad y^2 + t^2 = 4 ,

using y(0)=2y(0) = 2 to get 44. This is an implicit solution: a relation between yy and tt that holds along the curve, without yy isolated. The relation describes a full circle, but a solution must be a function of tt, so the solution through (0,2)(0, 2) is the upper half, the explicit solution

y=4−t2,−2<t<2.y = \sqrt{4 - t^2}, \qquad -2 \lt t \lt 2 .

The sign was chosen by the initial condition (y(0)=2>0y(0) = 2 \gt 0). The interval is open because at t=±2t = \pm 2 we have y=0y = 0 and the slope −t/y-t/y is undefined: the curve has a vertical tangent and cannot be continued as a function. The equation is perfectly smooth for y≠0y \neq 0. The solution ends because the curve runs into the line y=0y = 0.

Often the explicit form does not exist. For y′=t/(y+ey)y' = t/(y + e^y) separation gives (y+ey)dy=t dt\left(y + e^y\right)dy = t\,dt, hence

y22+ey=t22+C,\frac{y^2}{2} + e^y = \frac{t^2}{2} + C ,

and no algebra isolates yy. As with the circle, the slope is undefined where y+ey=0y + e^y = 0. The implicit relation is still the complete answer: for each tt you can find yy by root-finding, exactly as you would solve any nonlinear equation numerically.

Exponential growth, decay and the half-life

The simplest separable equation is y′=kyy' = ky. Following the recipe (the root of g(y)=yg(y) = y is y≡0y \equiv 0, noted):

∫dyy=∫k dt⟹ln⁡∣y∣=kt+c⟹y=±ec ekt.\int \frac{dy}{y} = \int k\,dt \quad\Longrightarrow\quad \ln|y| = kt + c \quad\Longrightarrow\quad y = \pm e^{c}\,e^{kt} .

The sign and ece^c merge into one arbitrary constant: y=Cekty = Ce^{kt} with C≠0C \neq 0. Notice that C=0C = 0 gives y≡0y \equiv 0, the solution we set aside. Here the general formula happens to recover it. Remember that this is luck, not a law; the next section shows a case where it does not.

With y(0)=y0y(0) = y_0 we get C=y0C = y_0. For k>0k \gt 0 this is growth, and the time for yy to double satisfies ekT=2e^{kT} = 2, so T=ln⁡2/kT = \ln 2 / k. For decay write k=−λk = -\lambda with λ>0\lambda \gt 0. The half-life solves e−λT1/2=12e^{-\lambda T_{1/2}} = \tfrac12:

T1/2=ln⁡2λ.T_{1/2} = \frac{\ln 2}{\lambda} .

It does not mention y0y_0: starting from any amount, the same time halves it. That is a property of exponentials alone, and it is why half-lives are quoted as constants of a substance.

Worked example (radiocarbon). Carbon-14 has T1/2=5730T_{1/2} = 5730 years, so λ=ln⁡2/5730=1.21×10−4\lambda = \ln 2/5730 = 1.21\times 10^{-4} per year. A sample retains 20% of its original carbon-14. Then e−λt=0.2e^{-\lambda t} = 0.2 and

t=ln⁡5λ=5730 ln⁡5ln⁡2≈13,300 years.t = \frac{\ln 5}{\lambda} = 5730\,\frac{\ln 5}{\ln 2} \approx 13{,}300 \text{ years}.

That is log⁡25=2.32\log_2 5 = 2.32 half-lives, a little more than two halvings (which would leave 25%).

Worked example (a cooling cup). Coffee at 90∘90^\circC sits in a 20∘20^\circC room: T′=−k(T−20)T' = -k(T - 20). Here g(T)=T−20g(T) = T - 20, whose root T≡20T \equiv 20 is the lost solution (room temperature, the steady state). Elsewhere ln⁡∣T−20∣=−kt+c\ln|T - 20| = -kt + c, so T=20+70e−ktT = 20 + 70e^{-kt}, and C=0C = 0 recovers T≡20T \equiv 20 again. If the coffee reaches 70∘70^\circC after 10 minutes then 50=70e−10k50 = 70e^{-10k}, so k=ln⁡(1.4)/10=0.0336k = \ln(1.4)/10 = 0.0336 per minute, and it reaches 40∘40^\circC when 20=70e−kt20 = 70e^{-kt}, i.e. t=ln⁡(3.5)/k=37t = \ln(3.5)/k = 37 minutes. The decaying quantity is the gap from equilibrium, and its half-life is ln⁡2/k=21\ln 2/k = 21 minutes.

The solutions you lose by dividing

Step 2 of the recipe divides by g(y)g(y). That is only legal where g≠0g \neq 0, and where g(c)=0g(c) = 0 the constant function y≡cy \equiv c satisfies the equation (both sides are zero), yet it never appears in ∫dy/g\int dy/g. It is a lost equilibrium solution, and forgetting it is the classic separation mistake.

Worked example. Solve y′=y2−1y' = y^2 - 1. Constant solutions: y≡1y \equiv 1 and y≡−1y \equiv -1. Elsewhere, partial fractions give 1y2−1=12(1y−1−1y+1)\frac{1}{y^2 - 1} = \frac12\left(\frac{1}{y-1} - \frac{1}{y+1}\right), so

12ln⁡∣y−1y+1∣=t+c⟹y−1y+1=Ke2t⟹y=1+Ke2t1−Ke2t.\frac12\ln\left|\frac{y-1}{y+1}\right| = t + c \quad\Longrightarrow\quad \frac{y-1}{y+1} = Ke^{2t}\quad\Longrightarrow\quad y = \frac{1 + Ke^{2t}}{1 - Ke^{2t}} .

Check the two lost solutions against the formula. K=0K = 0 gives y≡1y \equiv 1: recovered. But y≡−1y \equiv -1 would need K=∞K = \infty. No finite KK produces it, and the initial condition y(0)=−1y(0) = -1 makes K=(y0−1)/(y0+1)K = (y_0 - 1)/(y_0 + 1) a division by zero, which is the formula’s way of saying so. Someone who stops at the formula would claim that y(0)=−1y(0) = -1 has no solution; in fact its solution is the constant y≡−1y \equiv -1.

The lost solutions are not a footnote. They are the steady states of the system (terminal velocity above is one), and they organize everything else. Because distinct solutions never cross, a solution starting between or beyond equilibria stays in its band. Here:

  • y(0)=0y(0) = 0 gives K=−1K = -1 and y=−tanh⁡ty = -\tanh t, which slides down into −1-1 without reaching it.
  • y(0)=2y(0) = 2 gives K=1/3K = 1/3, and the denominator 1−e2t/31 - e^{2t}/3 vanishes at t=12ln⁡3=0.549t = \tfrac12\ln 3 = 0.549, where the solution blows up, as the next section explains. Above the line y=1y = 1 solutions are pushed upward; every solution below y=1y = 1 is drawn toward −1-1.
✅ Concept check

You solve $y' = t\,(y^2 - 9)$ by dividing both sides by $y^2 - 9$. Which constant solutions must you add back?

Domain of validity and finite-time blow-up

Now the equation that Euler could not see. Take y′=y2y' = y^2 with y(0)=y0≠0y(0) = y_0 \neq 0. The constant y≡0y \equiv 0 is set aside, then

∫dyy2=∫dt⟹−1y=t+C⟹y=y01−y0t,C=−1y0.\int \frac{dy}{y^2} = \int dt \quad\Longrightarrow\quad -\frac1y = t + C \quad\Longrightarrow\quad y = \frac{y_0}{1 - y_0 t}, \qquad C = -\frac{1}{y_0}.

For y0=0.5y_0 = 0.5 this is y=0.5/(1−0.5t)=1/(2−t)y = 0.5/(1 - 0.5t) = 1/(2 - t). The denominator vanishes at

t∗=1y0=2,t^* = \frac{1}{y_0} = 2 ,

and y→+∞y \to +\infty as t→2−t \to 2^-. There is the t=2t = 2 of the last two lessons. Euler’s 1.831.83 at t=2t = 2 was a number for a solution that does not exist.

Three points about what “valid” means here.

The interval belongs to the solution, not the equation. The right side y2y^2 is smooth everywhere, so nothing in ff announces trouble. The maximal interval for y0=0.5y_0 = 0.5 is (−∞,2)(-\infty, 2); for y0=2y_0 = 2 it is (−∞,0.5)(-\infty, 0.5); for y0<0y_0 \lt 0 the pole sits at the negative time 1/y01/y_0, so the solution lives on (1/y0,∞)(1/y_0, \infty) and decays; for y0=0y_0 = 0 (the lost solution) it is all of R\mathbb{R}. The same equation, four different answers to “how long does it last”.

The formula continues past t∗t^*, the solution does not. For t>2t \gt 2 the expression 1/(2−t)1/(2-t) is negative and satisfies the ODE on (2,∞)(2, \infty), but that is the solution through, say, (3,−1)(3, -1). It is not a continuation of ours, since the solution must pass through t=2t = 2 with a finite value and none exists there.

There is a mechanism. For y′=g(y)y' = g(y) with g>0g \gt 0, the time to travel from y0y_0 to ∞\infty is

∫y0∞dyg(y).\int_{y_0}^{\infty}\frac{dy}{g(y)} .

For g=y2g = y^2 this equals 1/y01/y_0, finite: the speed grows so much faster than yy that infinity is reached in finite time. For g=yg = y it diverges (the logarithm has no limit), so exponential growth never blows up. In general y′=ypy' = y^p blows up in finite time exactly when p>1p \gt 1; for p=3p = 3 the same calculation gives t∗=1/(2y02)t^* = 1/(2y_0^2). Finite-time blow-up is usually the model’s way of saying that a neglected effect, such as saturation, has become important.

Analytic against RK4

Here is the module’s standing habit, finally with something to compare against. The sim runs RK4 (or Euler) next to the exact solution for four separable equations and shades the region beyond the end of the exact domain.

◆ Interactive Exact solution against RK4 or Euler, with the end of the exact domain marked

Loading simulation…

Things to try, each of which the readout confirms:

  • y′=y2y' = y^2, y(0)=0.5y(0) = 0.5, RK4, h=0.25h = 0.25: the exact curve and RK4 agree closely on the approach (error 0.0280.028 at t=1.75t = 1.75, exact value 44), and RK4 stops after t=2t = 2 (its last finite value is 32.832.8). Compare Euler with h=0.5h = 0.5: it goes past t∗t^* and still reports finite values.
  • y′=−t/yy' = -t/y, y(0)=1y(0) = 1, RK4 with h=0.1h = 0.1: after the exact solution ends at t=1t = 1, RK4 keeps producing finite values. It hops from circle to circle: y2+t2y^2 + t^2 is 11 up to t=1t = 1, then 1.551.55, 1.771.77, 2.402.40, and finally 8.198.19 for the rest of the run, with no warning of any kind. The invariant caught it; the values alone would not.
  • y′=y2−1y' = y^2 - 1: put y(0)=1y(0) = 1 and y(0)=−1y(0) = -1 and watch the numerical solution sit exactly on the constant lines.
✅ Concept check

For $y' = y^2$ with $y(0) = 4$, at what time $t^*$ does the solution blow up? (Give a number.)

🐍 PythonPlayground: RK4 against the exact solution of y' = y^2 near the blow-up
⌘/Ctrl + Enter

With h=0.1h = 0.1 RK4 is at 9.9299.929 against the exact 1010 at t=1.9t = 1.9 (0.7% off), then reads 8282 at t=2t = 2 (no exact value exists) and 101210^{12} at t=2.1t = 2.1. Cutting hh tenfold moves the explosion only from t=2.1t = 2.1 to t=2.01t = 2.01 and the value at t=2t = 2 from 8282 to 820820, which is the right behaviour: a pole at t∗=2t^* = 2 cannot be resolved by any finite hh, but RK4 shows the explosion within a step or two of the truth, whereas Euler’s (last lesson) came at about t=4.5t = 4.5. Convergence is a statement about intervals [0,T][0, T] with T<t∗T \lt t^*.

Exercise: symbolic against numerical

Solve y′=t y2y' = t\,y^2, y(0)=1y(0) = 1 with sympy, then check it with RK4. By hand: −1/y=t2/2+C-1/y = t^2/2 + C, and C=−1C = -1, so y=1/(1−t2/2)y = 1/(1 - t^2/2), with a lost solution y≡0y \equiv 0 and a blow-up at t∗=2t^* = \sqrt 2. Your code should produce the same things.

🧪 Coding exerciseExercise: separate with sympy, then compare with RK4
⌘/Ctrl + Enter
✏️ Pen & paperSeparate and state the interval●○○

Solve y′=t e−yy' = t\,e^{-y}, y(0)=0y(0) = 0, explicitly, and give the interval on which the solution is valid.

Show worked solution

g(y)=e−yg(y) = e^{-y} is never zero, so no solutions are lost. Then ey dy=t dte^{y}\,dy = t\,dt, so ey=t2/2+Ce^y = t^2/2 + C. From y(0)=0y(0) = 0: 1=C1 = C. Therefore

y=ln⁡(1+t22).y = \ln\left(1 + \frac{t^2}{2}\right) .

The argument of the logarithm is at least 11 for every tt, so the solution is valid for all real tt.

✏️ Pen & paperHalf-life from a measurement●●○

A radioactive sample falls to 80% of its mass in 10 days. Find the decay constant, the half-life, and the time for it to fall to 10%.

Show worked solution

y=y0e−λty = y_0e^{-\lambda t} and e−10λ=0.8e^{-10\lambda} = 0.8, so λ=ln⁡(1.25)/10=0.02231\lambda = \ln(1.25)/10 = 0.02231 per day. The half-life is ln⁡2/λ=31.1\ln 2/\lambda = 31.1 days. For 10%: t=ln⁡10/λ=103.2t = \ln 10/\lambda = 103.2 days. (As a check, 103.2 days is 3.323.32 half-lives and 2−3.32=0.102^{-3.32} = 0.10.)

✏️ Pen & paperLost solutions and a blow-up time●●●

Consider y′=y2−4y' = y^2 - 4. (a) List the constant solutions. (b) Solve with y(0)=3y(0) = 3 and find when it blows up. (c) For y(0)=1y(0) = 1, find lim⁡t→∞y\lim_{t\to\infty} y, and say whether the general formula can produce the solution through y(0)=−2y(0) = -2.

Show worked solution

(a) y2−4=0y^2 - 4 = 0 gives y≡2y \equiv 2 and y≡−2y \equiv -2.

(b) 1y2−4=14(1y−2−1y+2)\frac{1}{y^2 - 4} = \frac14\left(\frac{1}{y-2} - \frac{1}{y+2}\right), so 14ln⁡∣y−2y+2∣=t+c\frac14\ln\left|\frac{y-2}{y+2}\right| = t + c, i.e. y−2y+2=Ke4t\frac{y-2}{y+2} = Ke^{4t} and

y=2(1+Ke4t)1−Ke4t.y = \frac{2\left(1 + Ke^{4t}\right)}{1 - Ke^{4t}} .

For y(0)=3y(0) = 3: K=1/5K = 1/5. The denominator vanishes at e4t=5e^{4t} = 5, i.e. t∗=14ln⁡5=0.402t^* = \tfrac14\ln 5 = 0.402.

(c) For y(0)=1y(0) = 1: K=−1/3K = -1/3, so Ke4t→−∞Ke^{4t} \to -\infty and y→−2y \to -2. The solution through y(0)=−2y(0) = -2 is the constant −2-2. The formula would need K=∞K = \infty, so it cannot produce it; the constant must be added by hand.

✅ Concept check

Which statement about the interval on which a solution of $y' = y^2$ exists is correct?

Where this goes next

Separation works when the right side factors. For y′=a(t)y+b(t)y' = a(t)y + b(t) it generally does not, but a different trick does: 3.4 Linear First-Order Equations & Integrating Factors. The question we assumed away in the lost-solution discussion, that exactly one solution passes through each point and for how long, is the subject of 3.7 Existence & Uniqueness, and the stability of the equilibria we found is taken up with autonomous equations in 3.8 Autonomous Equations, Equilibria & Stability. You can now check a solver: next lesson’s solutions will be tested the same way.

Key Takeaways

  • A separable equation y′=f(t)g(y)y' = f(t)g(y) is solved by ∫dy/g(y)=∫f dt+C\int dy/g(y) = \int f\,dt + C. The “fraction” step is justified by the chain rule: ddtH(y(t))=y′/g(y)\frac{d}{dt}H(y(t)) = y'/g(y), so H(y(t))H(y(t)) and F(t)F(t) differ by a constant.
  • A solution may stay implicit (y2+t2=4y^2 + t^2 = 4), and the explicit branch is chosen by the initial condition. Its interval stops where the curve reaches a vertical tangent.
  • Exponential growth and decay: y′=kyy' = ky gives y0ekty_0e^{kt}; the half-life ln⁡2/λ\ln 2/\lambda and doubling time ln⁡2/k\ln 2/k do not depend on y0y_0.
  • Dividing by g(y)g(y) discards the constant solutions g(c)=0g(c) = 0. Check whether the formula recovers them: for y′=y2−1y' = y^2 - 1, y≡1y \equiv 1 is recovered and y≡−1y \equiv -1 is not. They are the steady states.
  • The interval of validity belongs to the solution: y′=y2y' = y^2 from y0y_0 blows up at t∗=1/y0t^* = 1/y_0, which for y0=0.5y_0 = 0.5 is the t=2t = 2 that Euler sailed past.
  • y′=g(y)y' = g(y) reaches infinity in finite time when ∫∞dy/g\int^\infty dy/g converges: ypy^p with p>1p \gt 1, not p=1p = 1.
  • A closed form makes your solvers testable: compare against it, and compare the invariant H(y)−F(t)H(y) - F(t) along a run.
Finished this lesson?

Mark it complete to track your progress (saved in this browser). Reading every section does this for you.

Where this leads

These later lessons build directly on this one: