Integrating factors for non-exact equations: Foundations

Before you can use the parent note, you must own every piece of its language. We build each symbol from nothing, anchor it to a picture, and say why the topic needs it. Read top to bottom — each item leans on the one above.

1. Functions of two variables: $M (x, y)$ , $N (x, y)$ , $F (x, y)$

Picture a flat map (the $x$ – $y$ plane laid on the ground). At every point you stand, the function hands you a number — think height above sea level. So $F$ is a landscape: a hill sitting over the plane.

$x$ = how far east you stand (horizontal axis).
$y$ = how far north you stand (the other horizontal axis).
$F (x, y)$ = the height of the hill directly above you.

Why the topic needs it. The whole method hunts for a hidden height-function $F$ whose contour lines are the solutions of the ODE. And $M$ , $N$ are themselves two-variable functions — they take a value at every point of the plane.

2. Contour lines and " $F = C$ "

On a real hiking map, contour lines join places of equal height. Walk along a contour and you never go up or down.

The green loops in the figure are the curves $F = 1$ , $F = 2$ , $F = 3$ . The whole point of solving the ODE is to name these loops — the answer " $F (x, y) = C$ " means "the solution is whichever contour you started on."

Why the topic needs it. The boxed answers in the parent (e.g. $x^{3} y + \frac{1}{2} x^{2} y^{2} = C$ ) are exactly "this contour of $F$ ." No contours, no solution.

3. The differential pieces $d x$ , $d y$ and the form $M d x + N d y$

Look at the yellow tiny arrow $(d x, d y)$ at a point. The rule scores it: $M$ weights the eastward part, $N$ weights the northward part. Choosing a direction that scores $0$ keeps you on the level curve.

Why the topic needs it. The parent writes every first-order ODE in this differential form $M d x + N d y = 0$ . That single line is the object we test, multiply, and solve.

4. Partial derivatives $\frac{\partial F}{\partial x}$ , $F_{x}$ , $M_{y}$ , $N_{x}$

The word partial means "one direction at a time." Freeze $y$ like a constant, then differentiate the leftover one-variable function of $x$ as usual.

Blue slice: cut the hill with a wall running east; the tilt of that slice is $F_{x}$ .
Red slice: cut with a wall running north; its tilt is $F_{y}$ .

Why the topic needs it. The exactness test $M_{y} = N_{x}$ , the integrating-factor PDE, and the recovery of $F$ are all built from partial derivatives. This is the single most-used tool on the page.

5. The total differential $d F = F_{x} d x + F_{y} d y$

This is just "rise = slope × run," done in both directions and added. It is the bridge between the hill picture and the walking-rule picture.

Why the topic needs it. This is the definition of exact: an equation is exact exactly when its left side is somebody's total differential.

6. Why $M_{y} = N_{x}$ tests exactness (Clairaut)

If $M = F_{x}$ and $N = F_{y}$ , then $M_{y} = F_{x y}$ and $N_{x} = F_{y x}$ , and Clairaut forces these equal. So:

Why the topic needs it. This is the cheap check that tells us whether an $F$ even exists before we hunt for it — and the equation the integrating factor is designed to force true. (Vector-field readers: this is the curl-free / conservative condition.)

7. The exponential and the integral: $exp$ , $\int$ , and $\frac{μ ^{'}}{μ}$

Why the topic needs it. The two boxed formulas $μ (x) = exp \int \frac{M _{y} - N _{x}}{N} d x$ and $μ (y) = exp \int \frac{N _{x} - M _{y}}{M} d y$ are this idea. This same $e^{\int P d x}$ shows up in Linear first-order ODEs — it is the same trick.

F (x, y)

picture as?

A landscape / hill: a height above every point of the flat

x

–

y

plane.

What is a contour line, and why is the answer written

F = C

All points of equal height

F = C

; solution curves are contours, one per constant

C

What does the walking rule

M d x + N d y = 0

ask geometrically?

Move in a direction that changes the quantity by nothing — stay on a level curve.

How do you compute

M_{y}

Freeze

x

as a constant and differentiate

M

with respect to

y

Write the total differential of

F

d F = F_{x} d x + F_{y} d y

What makes an equation exact in terms of

F

Its left side equals a total differential:

M = F_{x}

and

N = F_{y}

Why does exactness reduce to

M_{y} = N_{x}

Because

M_{y} = F_{x y}

N_{x} = F_{y x}

, and Clairaut gives

F_{x y} = F_{y x}

Why does

exp

appear in the integrating factor?

Solving

μ^{'} / μ = g

needs the function whose

ln

-derivative is

g

; that is

μ = exp \int g

Why is the integrating factor never zero?

exp

is always positive, so multiplying by it keeps the same solution set.

Why does

ϕ (y)

appear when recovering

F

The

x

-integration "constant" can be any function of

y

, since such terms have zero

x

-slope.

Connections

Integrating factors for non-exact equations — the parent this page equips you for
Exact differential equations
Total differentials and potential functions
Mixed partial derivatives (Clairaut's theorem)
Linear first-order ODEs
Separable equations
Conservative vector fields

Foundations — Integrating factors for non-exact equations

1. Functions of two variables: $M (x, y)$ , $N (x, y)$ , $F (x, y)$

2. Contour lines and " $F = C$ "

3. The differential pieces $d x$ , $d y$ and the form $M d x + N d y$

4. Partial derivatives $\frac{\partial F}{\partial x}$ , $F_{x}$ , $M_{y}$ , $N_{x}$

5. The total differential $d F = F_{x} d x + F_{y} d y$

6. Why $M_{y} = N_{x}$ tests exactness (Clairaut)

7. The exponential and the integral: $exp$ , $\int$ , and $\frac{μ ^{'}}{μ}$

8. The recovery of $F$ by partial integration and $ϕ (y)$

How it all feeds the topic

Equipment checklist

Connections