These ask only: is it linear, and what are P and Q? No solving yet.
Recall Solution 1.1
(a) Linear.y and dxdy appear only to the first power, never multiplied together. Already standard (coefficient of dxdy is 1): P=3,Q=x2.
(b) NOT linear. The term y2 raises y to the second power — the definition forbids any power of y other than the first.
(c) Linear. But coefficient of dxdy is x, not 1. Divide by x: dxdy+x1y=xsinx, so P=x1,Q=xsinx.
(d) NOT linear.dxdy sits inside sin(⋅) — a function of the derivative. Linear means y and y′ appear bare, to the first power, never inside another function.
Recall Solution 1.2
(a)μ=e∫5dx=e5x.
(b)μ=e∫x2dx=e2ln∣x∣=∣x∣2=x2 (take x2 on our working domain).
(c) Here P=−tanx, so μ=e∫−tanxdx. Since ∫tanxdx=−ln∣cosx∣, we get ∫−tanxdx=ln∣cosx∣, hence μ=eln∣cosx∣=∣cosx∣=cosx (on an interval where cosx>0).
Standardize: already standard, P=2x,Q=x.
Mu:μ=e∫2xdx=ex2.
Multiply & collapse:dxd(ex2y)=xex2.
Integrate: the right side is ∫xex2dx. Why substitution here? Because the derivative of the exponent x2 is 2x, and we have an x outside — a classic "inner-derivative-is-present" signal. Let u=x2, du=2xdx:
∫xex2dx=21∫eudu=21ex2+C.
So ex2y=21ex2+C.
Divide:y=21+Ce−x2.
Now the equation is disguised: you must standardize, or the integral needs a technique.
Recall Solution 3.1
Standardize: divide by x (coefficient of dxdy must be 1): dxdy−x2y=x3. So P=−x2,Q=x3.
Mu:μ=e∫−x2dx=e−2lnx=x−2=x21.
Multiply & collapse:dxd(x−2y)=x−2⋅x3=x.
Integrate:x−2y=∫xdx=2x2+C.
Divide: multiply by x2: y=2x4+Cx2.
Recall Solution 3.2
Standardize:P=1,Q=x.
Mu:μ=e∫1dx=ex.
Multiply & collapse:dxd(exy)=xex.
Integrate by parts:Why parts? We have a product of "polynomial × exponential" and no chain-rule inner derivative is present, so substitution won't help — parts is the tool that trades a hard product for an easier one. With u=x,dv=exdx:
∫xexdx=xex−∫exdx=xex−ex+C.
So exy=xex−ex+C.
Divide:y=x−1+Ce−x.
Apply IC last:y(0)=2⇒2=0−1+C⇒C=3. Answer:y=x−1+3e−x.
Recall Solution 3.3
Standardize: divide by cosx (nonzero on this interval): dxdy+tanxy=secx. So P=tanx,Q=secx.
Mu:μ=e∫tanxdx=e−ln∣cosx∣=cosx1=secx (since cosx>0 here).
Multiply & collapse:dxd(secxy)=secx⋅secx=sec2x.
Integrate:secxy=∫sec2xdx=tanx+C.
Divide:y=secxtanx+C=sinx+Ccosx (using secxtanx=sinx and secx1=cosx).
Combine the method with substitutions, other ODE types, or geometry.
Recall Solution 4.1
This is a Bernoulli equation with n=2 — not linear as written (the xy2 term multiplies powers). Substitutev=y1−n=y−1, so dxdv=−y−2dxdy.
Divide the ODE by y2: y−2dxdy+y−1=x. In terms of v: −dxdv+v=x, i.e.
dxdv−v=−x.
Now it's linear in v: P=−1,Q=−x. μ=e∫−1dx=e−x.
dxd(e−xv)=−xe−x. Integrate by parts: ∫−xe−xdx=xe−x+e−x+C.
So e−xv=xe−x+e−x+C⇒v=x+1+Cex.
Back-substitute v=1/y: y1=x+1+Cex, so y=x+1+Cex1.
Recall Solution 4.2
μ=ex throughout. On 0≤x≤1 (f=1): dxd(exy)=ex⇒exy=ex+C1⇒y=1+C1e−x. IC y(0)=0: 0=1+C1⇒C1=−1, so
y=1−e−x(0≤x≤1).
At x=1: y(1)=1−e−1.
On x>1 (f=0): dxd(exy)=0⇒exy=C2⇒y=C2e−x. Continuity at x=1: C2e−1=1−e−1⇒C2=e−1, so
y=(e−1)e−x(x>1).
As x→∞, y→0: with the driving turned off, the solution decays.
Recall Solution 4.3
Rewrite: dxdy−x1y=1, so P=−x1,Q=1.
μ=e∫−x1dx=e−lnx=x1.
dxd(xy)=x1⋅1=x1. Integrate: xy=lnx+C.
So y=xlnx+Cx. IC (1,0): 0=1⋅0+C⋅1⇒C=0. Answer:y=xlnx.
Shape: at x=1, y=0; the red curve dips slightly negative just right of x=1 (since lnx<0 for x<1... on x>1 it climbs), touches a minimum where dxdy=lnx+1=0, i.e. x=e−1, then rises steadily — a gently curving line pinned through the marked point.
Structural insight, parameters, and existence subtleties.
Recall Solution 5.1
Case k=−1:μ=ex, dxd(exy)=exe−x=1. Integrate: exy=x+C, so y=(x+C)e−x. Notice the x out front — the forcing e−x matches the decaying mode e−x, so integrating a constant produces a linear factor.
Case k=−1:dxd(exy)=e(k+1)x, integrate: exy=k+11e(k+1)x+C, so y=k+11ekx+Ce−x. No stray x — the forcing and the natural mode differ, so a plain exponential particular solution works.
Insight: the x-multiplier appears exactly when Q's exponential rate equals the homogeneous rate −P. (This is the ODE analogue of resonance.)
Recall Solution 5.2
Let μ=e∫P, so μ′=μP (from the derivation's condition). Write μy=∫μQdx+C. Differentiate both sides with the product rule:
dxd(μy)μ′y+μy′=μQ.
Replace μ′=μP: μPy+μy′=μQ. Divide by μ (never zero since it's an exponential): y′+Py=Q. ✓
Exact-equation view: multiplying y′+Py−Q=0 by μ makes the left side the exact differentialdxd(μy) — the integrating factor is precisely what turns a non-exact form into an exact one.
Recall Solution 5.3
(a) Standardize (for x=0): dxdy+x1y=x1, μ=x. Then dxd(xy)=x⋅x1=1, so xy=x+C⇒y=1+xC.
(b) The standard form has P=x1 and Q=x1, both undefined at x=0 — the existence-uniqueness theorem needs P,Q continuous, which fails there. Looking at y=1+xC: as x→0, the term xC blows up unless C=0. So the only solution that stays finite at x=0 is y=1, giving y(0)=1. Therefore an initial condition y(0)=y0 is admissible only if y0=1; any other value has no solution reaching the origin. (Notice the original equation at x=0 reads 0+y=1, forcing y(0)=1 — consistent.)