Antiderivative — definition, family of solutions (+C): Visual walkthrough

4.2.1 · D2 · Maths › Calculus II — Integration › Antiderivative — definition, family of solutions (+C)

Shuru karne se pehle, teen simple-words ke anchors taaki koi symbol bina samjhe na aaye:

Step 1 — Ek slope-report bas har point par ek arrow hai

KYA. Hum ek slope-report $f (x)$ lete hain aur use graph ki tarah nahi, balki chote tilted arrows ke ek field ki tarah draw karte hain — plane ke har point par ek short line segment, $f$ jo slope demand karta hai us par tilted.

KYUN. Yeh antiderivative ko jo diya jaata hai uski honest picture hai. Humein heights nahi batai gayi; humein tilts batai gayi hain. Pehle tilts draw karna humein us height ko secretly assume karne se rokta hai jo humein kabhi di hi nahi gayi.

PICTURE. Figure dekho: har grid point par ek short pastel dash baitha hai. Uski steepness $f$ se fix hai; vertical position abhi tak kisi bhi cheez se fix nahi hai — yahi freedom poori kahani hai.

Figure — Antiderivative — definition, family of solutions (+C)

$f (x)$ — woh slope jo horizontal position $x$ par arrow ke paas hona chahiye.
Har dash — ek instruction: "yahan se guzarne wala curve itna steep hona chahiye."

Step 2 — Arrows ke through ek curve thread karna

KYA. Ab hum ek single curve $F$ trace karte hain jo har us arrow ke tangent (flush) rehta hai jiske paas se guzarta hai. Jahan bhi jaaye, uski ruler-tilt local arrow ke barabar ho.

KYUN. Ek curve jo field ke har jagah tangent hai, exactly woh curve hai jiska slope-report $f$ hai — matlab $F^{'} (x) = f (x)$ . Yeh antiderivative ki definition hai, ab likhi nahi balki draw ki gayi.

PICTURE. Bold lavender curve arrows ko hug kar raha hai. Notice karo ki humein isse kahan se start karna hai choose karna pada — humne isse ek height par drop kiya aur arrows ne baki steering ki.

$F^{'} (x) = f (x)$

$F^{'} (x)$ — $x$ par hamare bold curve ki tilt (uska ruler).
$f (x)$ — woh tilt jo arrow field ne $x$ par demand ki thi.
$=$ — "curve field se match karta hai": yahi $F$ ko antiderivative banata hai.

Step 3 — Ise upar slide karo: slope kabhi nahi badlta

KYA. Bold curve lo aur ise ek fixed amount $C$ se seedha upar shift karo. Nayi curve ko $G (x) = F (x) + C$ bulao. Arrows ke against uski tilt check karo.

KYUN. Vertically sliding karne se height move hoti hai lekin koi tilt saath nahi aati — ek utha hua pahaad same shape ka hota hai. Ise symbols mein dekhne ke liye hum ek hi fact use karte hain ki ek flat, constant line ki slope $0$ hoti hai: $G^{'} (x) = \frac{d}{d x} (F (x) + C) = F^{'} (x) + = 0 \frac{d}{d x} (C) = f (x) .$

PICTURE. Do curves, coral lavender ke upar, vertical gap $C$ se offset. Marked $x$ par unke chote rulers parallel hain — same tilt, alag height. Raised copy abhi bhi same arrows ko hug karti hai.

$C$ — ek constant, fixed vertical gap (jitna hum slide kiye).
$\frac{d}{d x} (C) = 0$ — ek horizontal line ki zero slope hoti hai, toh shifting tilt mein kuch add nahi karti.
$G^{'} (x) = f (x)$ — shifted curve bhi ek valid antiderivative hai.

Toh kam se kam utne antiderivatives hain jitne $C$ ke choices hain — infinitely many.

Step 4 — Kya koi sneaky aur curves hain, sirf vertical slides nahi?

KYA. Maano koi claim karta hai ki ek second antiderivative $G$ hai jo $F$ ki plain vertical copy nahi hai. Hum us claim ko subtract karke test karte hain: gap function $H (x) = G (x) - F (x)$ define karo.

KYUN. Agar $G$ aur $F$ dono field se match karte hain, toh unka difference bilkul bhi koi tilt carry nahi karna chahiye. Subtract karne se shared slope cancel ho jaata hai aur jo bhi "extra shape" $G$ mein secretly ho sakti hai woh isolate ho jaati hai: $H^{'} (x) = G^{'} (x) - F^{'} (x) = f (x) - f (x) = 0 for all x .$

PICTURE. Top panel: do candidate curves. Bottom panel: unka gap $H$ , ek green track ki tarah draw kiya gaya jiska ruler har jagah flat hai — har $x$ par slope $0$ .

$H (x) = G (x) - F (x)$ — har $x$ par do candidates ke beech vertical distance.
$H^{'} (x) = 0$ — us distance ki zero slope hai; yeh kabhi tilt nahi karta.

Step 5 — Har jagah zero slope ek flat line force karta hai (MVT step)

KYA. Hum prove karte hain ki gap $H$ secretly rise ya fall nahi kar sakta — yeh perfectly flat, constant line hona chahiye.

KYUN. Yeh ek jagah hai jahan humein ek named theorem ki zaroorat hai. Mean Value Theorem kehta hai: koi bhi do points ke beech, ek smooth curve ko kahin na kahin unke beech average steepness par ek ruler tilted hona chahiye. Lekin $H$ par har ruler $0$ read karta hai. Toh kisi bhi do points ke beech average steepness $0$ hai — matlab do points same height par hain. Har pair ke liye same height ⇒ poori cheez ek level line hai.

PICTURE. $H$ par do points choose karo. Unke beech dashed "average-slope" chord draw ki gayi hai — aur kyunki saare rulers flat hain, woh chord horizontal hai, jo right point ko left point ki height par force karta hai. Pair ko kahin bhi slide karo: hamesha horizontal.

$H^{'} (x) = 0 on interval I ⟹ H (x) = C .$

Dashed chord — woh average slope jo MVT promise karta hai kahin achieve hogi.
Horizontal chord — average slope $0$ , toh koi height change possible nahi.
$H (x) = C$ — gap ek single constant hai.

Step 4 aur Step 5 ko milao: $G (x) - F (x) = C ⟹ G (x) = F (x) + C .$ Koi bhi "aur" antiderivative har hal mein ek vertical slide hi tha. Family exactly $F + C$ hai — na zyada, na kam.

Step 6 — Degenerate case: agar domain mein koi hole ho toh?

KYA. Upar sab kuch quietly ek connected interval $I$ assume kar raha tha. Ab ground break karo: ek point remove karo, toh domain do separate stretches ban jaati hai (classic case hai $f (x) = \frac{1}{x}$ , $x = 0$ par undefined).

KYUN. Step 5 ka MVT do points compare karta hai sirf tab jab tum unke beech domain chhode bina travel kar sako. Ek gap ke across koi chord draw nahi ho sakti — toh left piece ka constant aur right piece ka constant agree karna zaruri nahi. Har connected island ko apna khud ka $C$ milta hai.

PICTURE. Ek domain jo ek shaded forbidden strip se split hai. Left island: ek curve $C_{1}$ se raised. Right island: ek independently raised curve $C_{2}$ se. Dono apne turf par same arrows ko hug karte hain; koi cheez unki heights ko tie nahi karti.

$\int \frac{1}{x} d x = {ln (- x) + C_{1}, ln (x) + C_{2}, x < 0 x > 0.$

Shaded strip — woh hole ( $x = 0$ ), jahan koi curve aur koi chord cross nahi kar sakti.
$C_{1}, C_{2}$ — do independent constants, ek per island.
$ln (- x)$ / $ln (x)$ — har side par antiderivative shape; Logarithmic & Exponential Integrals dekho ki log exactly power-rule exception $n = - 1$ par kyun aata hai.

Step 7 — Kaunsa curve choose karna hai: ek single point $C$ pin kar deta hai

KYA. Infinite stack mein se, demand karo ki curve ek known point $F (x_{0}) = y_{0}$ se guzre.

KYUN. Stack mein exactly ek free knob hai, $C$ (Steps 3–5). Ek equation ek unknown determine karta hai, toh ek single point ek single member select karta hai. Yahi poora idea hai ek initial value problem ke peeche.

PICTURE. Parallel curves ka pastel stack, ek coral point $(x_{0}, y_{0})$ marked. Sirf ek curve us dot se thread karti hai; baki dim hain.

Ek-picture summary

KYA. Ek figure poore walkthrough ko compress karta hai: arrow field (Step 1), parallel curves ka infinite stack jo sab iske tangent hain (Steps 2–5), vertical gap $C$ unme se do ke beech labelled (Step 3), aur ek highlighted member jo ek point se chosen hai (Step 7).

Arrow field — diya hua slope-report $f$ .
Stack — family $F (x) + C$ ; har curve har $x$ par same tilt share karti hai.
Gap $C$ — ek aur sirf ek freedom, Steps 4–5 se proven complete.
Coral dot — ek initial condition jo ek single curve select karti hai.

Recall Feynman retelling — poora walkthrough simple words mein

Koi tumhe har step par ek pahaad ki tilt batata hai lekin woh height kabhi nahi batata jahan se tumne shuru kiya. Pehle tum un tilts ko field mein tiny dashes ki tarah scatter karte ho (Step 1). Tum ek curve thread karte ho jo exactly unke saath lean karti hai (Step 2). Phir tum notice karte ho: poore curve ko seedha upar uthao aur woh abhi bhi same tarah lean karti hai — kyunki ek flat lift koi tilt add nahi karti (Step 3). Kya koi bilkul alag shape ka curve bhi fit ho sakta hai? Use apna subtract karo; leftover ki har jagah zero tilt hai (Step 4), aur koi cheez jis ki ek stretch mein har jagah zero tilt ho woh upar ya neeche nahi ja sakti — woh dead flat hai, ek constant (Step 5, Mean Value Theorem). Toh "alag" curve actually tumhara hi tha kisi fixed $C$ se upar slide kiya hua. Ek hi trap hai: agar ground mein ek aisa gap ho jis par tum walk nahi kar sakte, toh do sides alag heights par float kar sakti hain, ek $C$ each (Step 6). Finally, mujhe bas ek jagah ki height batao aur main infinite stack mein se exact curve pick kar leta hoon (Step 7). Woh unknown starting height, aur kuch nahi, wahi $+ C$ hai.

Connections

Parent topic — Antiderivative & the family +C
Mean Value Theorem — Step 5 ka engine.
Fundamental Theorem of Calculus — jahan ek chosen curve areas se milti hai.
Power Rule (Integration) & Logarithmic & Exponential Integrals — Step 6 mein $n = - 1$ ka hole.
Differential Equations — Initial Value Problems — Step 7 formalized.
Definite Integral — jahan $+ C$ harmlessly cancel ho jaata hai.