Visual walkthrough — Net change theorem

4.2.5 · D2 · Maths › Calculus II — Integration › Net change theorem

---

Characters ka parichay (Step 1 se pehle)

Humein do cheezein aur ek seedhi picture chahiye, warna neeche kuch bhi samajh nahi aayega.

Poora theorem $F$ ki ek picture (ek curve jo upar neeche jaati hai) ko $F^{\prime }$ ki ek picture (ek alag curve ke neeche area) se jodta hai. Dono ko ek saath side by side dekho:

Left board par, blue curve $F$ hai; $a$ se $b$ tak total vertical rise — pink bracket $F(b)-F(a)$ — woh net change hai jo hum chahte hain. Right board par $F^{\prime }$ hai, us same curve ki slope-o-meter reading. Jo claim hum prove karenge woh yeh hai ki right par yellow area left par pink height ke barabar hai. Alag pictures, same number.

---

Step 1 — Interval ko thin strips mein slice karo

KYA. $[a,b]$ ko $n$ equal chhote pieces mein kaato. Cut points mark karo $a=x_0<x_1<x_2<\cdots <x_n=b.$ Yahan $x_0$ bas $a$ ka doosra naam hai, $x_n$ doosra naam $b$ ka, aur har $x_i$ $i$-vaan fencepost hai. Har strip ki same width hai $\Delta x=\underset{\text{total length}÷\text{number of pieces}}{\underbrace{\frac{b-a}{n}}}.$ Symbol $\Delta$ (Greek "delta") ka matlab hai "mein ek change," toh $\Delta x$ literally padha jaata hai "x mein ek chota change" — strip ki width.

KYUN. Poore lambe interval par change ke baare mein sochna mushkil hai — $F^{\prime }$ wiggle karta hai. Lekin ek strip par jo itni patli hai ki $F$ muskil se bend ho, change aasaan hai: woh almost sirf (slope) $\times$ (width) hai. Slicing ek mushkil problem ko $n$ aasaan waalon se badal deti hai.

PICTURE.

Board dikhata hai interval ko $n$ equal chalk strips mein kata hua, har ek $\Delta x$ width ka. Fenceposts $x_0,\dots ,x_n$ dekho: $n$ gaps banane ke liye $n+1$ posts hain. Yoh off-by-one classic trap hai — gaps ginne hain, posts nahi.

---

Step 2 — Total change ek telescoping sum hai

KYA. Poora rise $F(b)-F(a)$ bas har strip ke apne chhote rise ka sum hai: $F(b)-F(a)={\sum }_{i=1}^n\underset{\text{rise across strip }i}{\underbrace{[F(x_i)-F(x_{i-1})]}}.$ $\sum$ (Greek "sigma") ka matlab hai "inhe jodo," aur $i$ $1$ se $n$ tak chalta hai, ek baar har strip ke liye. Term $F(x_i)-F(x_{i-1})$ batata hai ki sirf strip number $i$ par $F$ kitna chada.

KYUN. Sum ko longhand mein likho aur dekho ki middle values kaise annihilate ho jaati hain: $(F(x_1)-F(x_0))+(F(x_2)-\cancel{F(x_1)})+(F(x_3)-\cancel{F(x_2)})+\cdots +(F(x_n)-\cancel{F(x_{n-1})}).$ Har interior height ek baar $+$ ke saath aur ek baar $-$ ke saath aata hai, toh cancel ho jaata hai. Sirf baahri survivors $F(x_n)=F(b)$ aur $-F(x_0)=-F(a)$ bachte hain. Yeh "beech se sab cancel ho jaata hai" pattern ek telescoping sum hai — ek collapsing telescope ki tarah. Yeh exact hai, approximation nahi.

PICTURE.

Har pink vertical bar ek strip ka rise $F(x_i)-F(x_{i-1})$ hai. Unhe staircase par head to tail stack karo aur poori staircase ki total height exactly pink bracket $F(b)-F(a)$ hai — kuch lost nahi, kuch double-count nahi.

---

Step 3 — Har chote rise ko slope × width se badlo (Mean Value Theorem)

KYA. Ek akele thin strip ke liye, uske rise ko ek slope times uski width se replace karo. Mean Value Theorem guarantee karta hai ki strip $i$ ke andar koi koi point $x_i^{\ast }$ hai jahan curve ka instantaneous slope strip ke average slope ke barabar hai: $F(x_i)-F(x_{i-1})=\underset{\text{slope at that point}}{\underbrace{F^{\prime }(x_i^{\ast })}}\cdot \underset{\text{strip width}}{\underbrace{\Delta x}}.$ $x_i^{\ast }$ par star bas tag karta hai "strip $i$ ke andar special point." Rearrange karo toh yeh kehta hai $\frac{\text{rise}}{\text{run}}=F^{\prime }(x_i^{\ast })$ — ek slope, by definition.

KYUN. Hum kahaani mein rate $F^{\prime }$ laana chahte hain, kyunki $F^{\prime }$ wahi hai jiske baare mein theorem hai. Mean Value Theorem bridge hai: woh ek aisa rise jo hum compute nahi kar sakte use ek rate mein convert karta hai (jo hum jaante hain) times ek width (jo humne choose ki). Yeh kaam karta hai kyunki ek strip par curve ko, kam se kam ek baar, apne hi start-to-end average jaisa steep hona chahiye — average slope ko match kiye bina tum bottom se top nahi ja sakte.

PICTURE.

Ek strip mein zoom in karo. Pink chord dono ends ko connect karta hai; yellow tangent line, $x_i^{\ast }$ par drawn, uske parallel hai. Same slope. Toh true rise (chord ka vertical drop) $F^{\prime }(x_i^{\ast })\Delta x$ ke barabar hai — height $F^{\prime }(x_i^{\ast })$ aur width $\Delta x$ ka ek genuine rectangle.

---

Step 4 — Rectangles ka ek Riemann sum pehchaano

KYA. Step 2 ke sum mein Step 3 ka replacement substitute karo: $F(b)-F(a)={\sum }_{i=1}^n\underset{\text{rectangle height}}{\underbrace{F^{\prime }(x_i^{\ast })}}\underset{\text{rectangle width}}{\underbrace{\Delta x}}.$ Har term $F^{\prime }(x_i^{\ast })\Delta x$ ek rectangle ka area hai: height $F^{\prime }(x_i^{\ast })$, width $\Delta x$.

KYUN. Yeh exactly ek Riemann sum ki shape hai — $F^{\prime }$ ke graph ke neeche baithe patle rectangles ka dhera. Humne ise purpose se engineer kiya: left side abhi bhi $F$ ka exact net change hai, aur ab right side clearly $F^{\prime }$ ke neeche area ka ek approximation hai.

PICTURE.

Yellow rectangles $F^{\prime }$ curve ke neeche march karte hain. Unka total area Step 1 ke left board ki pink net-change height ke barabar hai — strips ki kisi bhi number ke liye, kyunki Step 3 strip by strip exact tha.

---

Step 5 — Limit lo: rectangles integral ban jaate hain

KYA. Strips ko infinitely thin hone do, $n\to \infty$ (toh $\Delta x\to 0$). Rectangle-sum sharpen hokar exact area ban jaata hai, jo wahi hai jiska matlab integral symbol hai: $F(b)-F(a)={\lim }_{n\to \infty }{\sum }_{i=1}^n{F}^{\prime }(x_i^{\ast })\Delta x={\int }_a^b{F}^{\prime }(x)dx.$ Lamba $\int$ "sum" ke liye ek stretched "S" hai; $dx$ width $\Delta x$ hai jo ek whisker tak shrink ho gayi. Right se left padho: $[a,b]$ par (rate $\times$ tiny width) ka sum.

KYUN. Jaise strips patli hoti hain, do cheezein saath mein khatam hoti hain: har strip ke andar curve ka bending, aur rise ko slope $\times$ width se approximate karne mein koi bhi error. Limit mein wobbly rectangle-tops smoothly $F^{\prime }$ par settle ho jaate hain, aur sum wahi hai area. Left side kabhi hili nahi — woh poore time $F(b)-F(a)$ thi — toh dono equal hain.

PICTURE.

Teen boards: coarse rectangles, finer rectangles, aur smooth filled area. Jagged top $F^{\prime }$ curve par settle ho jaata hai; yellow area ab ${\int }_a^b{F}^{\prime }(x)dx$ hai, pink height ke bilkul barabar. $■$

---

Step 6 — Edge case: jab rate negative ho jaata hai

KYA. Upar kuch bhi yeh assume nahi karta tha ki $F^{\prime }\ge 0$. Agar $[a,b]$ ke kisi hisse par $F^{\prime }(x)<0$ hai, toh us strip ke rectangle ki negative height hai, toh woh negative area contribute karta hai — aur wahan $F$ neeche jaata hai.

KYUN. Negative rate ka matlab hai stuff decrease ho raha hai (paani drain ho raha hai, car reverse kar rahi hai). Proof abhi bhi line for line hold karta hai: Step 3 ka slope $F^{\prime }(x_i^{\ast })$ bas ek negative number hai, toh rise negative hai. Sum phir ups ko downs ke saath net karta hai. Yahi reason hai ki theorem net change deta hai, total distance nahi.

PICTURE.

$F^{\prime }$ curve beech wale stretch par axis ke neeche dip karta hai. Axis ke upar rectangles yellow hain (positive, $F$ upar ja raha hai); axis ke neeche rectangles pink hain (negative, $F$ gir raha hai). Integral unhe signed jodta hai: yellow area minus pink area. Woh signed total exactly $F(b)-F(a)$ hai — blue $F$ curve ko dekho upar jaata, phir girta, phir uthta, pink net height par khatam hota.

---

Step 7 — Degenerate cases: check karo machine tooti nahi

KYA & KYUN. Teen limiting inputs, har ek ek sanity test.

1. Empty interval, $a=b$. Zero total width ke zero strips: ${\int }_a^a{F}^{\prime }dx=F(a)-F(a)=0$. Koi time nahi guzra, koi change nahi. ✓
2. Constant quantity, $F^{\prime }\equiv 0$. Har rectangle ki height $0$ hai, toh area $0$ hai aur $F(b)-F(a)=0$. Flat stuff kabhi change nahi hota. ✓
3. Constant rate, $F^{\prime }\equiv c$. Har rectangle ki same height $c$ hai; total area $=c\cdot (b-a)$, ek saada rectangle. Aur sach mein net change $=c(b-a)$ — fixed time ke liye steady rate. ✓

PICTURE.

Teen mini-boards jo collapsed strip ($a=b$), flat-zero rate, aur constant rate ka ek single fat rectangle dikhate hain. Har case mein yellow area aur pink height agree karte hain — theorem gracefully degrade karta hai.

---

Ek-picture summary

Arrow chain padho: $F$ ki pink rise (left) chote rises ke staircase ke barabar hai (telescoping), yellow rectangles ke sum ke barabar hai (MVT), equals — limit mein — $F^{\prime }$ ke neeche yellow area (right). Ek number, chaar tareekon se dekha.

Recall Feynman: poora walkthrough ek 12-saal ke bachche ko batao

Ek pahaad socho. Pahaad ki sea level se upar height $F$ hai; jahan tum khade ho wahan woh kitna steeply dhalaan hai woh $F^{\prime }$ hai. Main jaanna chahta hoon ki door wala end $b$ kareebi end $a$ se kitna upar hai — woh net change hai.

Main walk ko chhote chhote steps mein slice karta hoon (Step 1). Har chhote step par zameen muskil se bend hoti hai, toh climb ka woh chota sa hissa bas "yahan kitna steep hai" times "step kitna lamba hai" hai (Step 3 — Mean Value Theorem promise karta hai ki har step mein ek jagah hai jahan steepness bilkul sahi hai). Agar main un saare chhote climbs ko stack karta hoon, beech wale cancel ho jaate hain aur main sirf "top minus bottom" ke saath reh jaata hoon (Step 2 — telescoping). Har chota climb ek patla rectangle hai, height = steepness, width = step length, toh poora climb steepness graph ke neeche rectangles ka dhera hai (Step 4). Steps ko infinitely small banao aur woh dhera smoothly steepness curve ke neeche area ban jaata hai — integral (Step 5).

Agar pahaad kahin downhill jaata hai, us step ka climb negative hai aur subtract ho jaata hai, toh mujhe net height gained milti hai, upar minus neeche (Step 6). Aur agar main bilkul nahi chalta, ya zameen flat hai, ya woh har jagah same dhalaan hai, machine phir bhi obvious answer deti hai (Step 7). Bas itna hi hai: steepness-times-tiny-width jodo aur tum height mein total change recover kar lete ho.

---

Connections

• Net change theorem — woh parent result jise yeh pictures prove karti hain.
• Fundamental Theorem of Calculus — same boxed equation, abstractly stated.
• Definite Integral as a Riemann Sum — Steps 4–5 yahi definition hain.
• Mean Value Theorem — Step 3 ka engine.
• Telescoping Sums — Step 2 ki cancellation.
• Displacement vs Distance — Step 6 ka net-vs-total split.
• Marginal Cost and Revenue — jahan signed area economics se milta hai.