4.1.27 · Maths › Calculus I — Limits & Derivatives
Intuition Badi picture (WHY ye matter karta hai)
Agar aap 1 ghante mein 100 km drive karte ho, toh aapki average speed 100 km/h hai. MVT kehta hai: kisi ek instant par aapki instantaneous speed (speedometer reading) exactly 100 km/h thi. Ye theorem guarantee karta hai ki average slope interval ke andar kisi jagah derivative dwara achieve hoti hai. Rolle's theorem woh special case hai jahan average slope 0 hoti hai.
Definition Rolle's Theorem
Maano f ek aisi function hai ki:
f closed interval [ a , b ] par continuous hai,
f open interval ( a , b ) par differentiable hai,
f ( a ) = f ( b ) .
Toh kam se kam ek aisa point c ∈ ( a , b ) exist karta hai jahan == f ′ ( c ) = 0 == .
Intuition YE kyon sach hona chahiye
Agar ek smooth curve same height par shuru aur khatam hoti hai, toh woh bina neeche aaye upar nahi ja sakti (ya bina upar aaye neeche nahi ja sakti). Turning point par — sabse oonchi peak ya sabse nichli valley par — tangent flat hoti hai, yaani f ′ ( c ) = 0 . Agar function har jagah perfectly flat hai, toh f ′ = 0 bhi har jagah hai.
Worked example Proof — Extreme Value Theorem se banaya gaya
Step 1 — Continuity use karke extremes paao.
Kyunki f closed bounded interval [ a , b ] par continuous hai, Extreme Value Theorem kehta hai ki f maximum value M aur minimum value m [ a , b ] par kahi na kahi attain karti hai.
Ye step kyon? Hume ek guaranteed "top" ya "bottom" point chahiye taaki flat tangent mil sake.
Step 2 — Do cases.
Case A: M = m . Tab f constant hai, isliye f ′ ( x ) = 0 har x ke liye; koi bhi c chuno.
Kyon? Ek flat line ki slope har jagah zero hoti hai.
Case B: M = m . Kyunki f ( a ) = f ( b ) , M , m mein se kam se kam ek interior point c ∈ ( a , b ) par attain hota hai (endpoints same value share karte hain, isliye woh dono strict max aur strict min nahi ho sakte).
Kyon? Agar dono extremes sirf endpoints par hote, toh M = f ( a ) = f ( b ) = m hota, jo M = m ka contradiction hai.
Step 3 — Interior extremum ⇒ derivative zero (Fermat's Theorem).
Maano c ek interior maximum hai. Chote h > 0 ke liye, f ( c + h ) ≤ f ( c ) , isliye
h f ( c + h ) − f ( c ) ≤ 0 ⇒ f ′ ( c ) = lim h → 0 + ≤ 0.
Chote h < 0 ke liye, f ( c + h ) ≤ f ( c ) lekin h < 0 sign flip karta hai:
h f ( c + h ) − f ( c ) ≥ 0 ⇒ f ′ ( c ) = lim h → 0 − ≥ 0.
Kyunki f differentiable hai, dono one-sided limits f ′ ( c ) ke barabar hain, isliye f ′ ( c ) ≤ 0 aur f ′ ( c ) ≥ 0 , jo f ′ ( c ) = 0 force karta hai. ■
Ye step kyon? Derivative dono sides se same limit hai; ise ≤ 0 aur ≥ 0 ke beech squeeze karna ise exactly 0 par pin kar deta hai.
Definition Mean Value Theorem
Maano f hai:
[ a , b ] par continuous , aur
( a , b ) par differentiable .
Toh ek aisa c ∈ ( a , b ) exist karta hai ki
f ′ ( c ) = b − a f ( b ) − f ( a ) .
Intuition Geometrically YE kya kehta hai
Right side b − a f ( b ) − f ( a ) endpoints ko join karne wali secant line ki slope hai. MVT kehta hai ki andar kahin ek tangent us secant ke parallel hai.
Worked example "Graph ko tilt karo" wali trick
Step 1 — Helper function banao. Define karo
g ( x ) = f ( x ) − [ f ( a ) + b − a f ( b ) − f ( a ) ( x − a ) ] .
Bracket exactly secant line L ( x ) hai. Isliye g ( x ) = f ( x ) − L ( x ) = curve aur secant ke beech vertical gap .
Ye step kyon? Hum secant subtract karte hain taaki dono endpoints same height par aa jaayein — phir Rolle apply hota hai.
Step 2 — g ke liye teen Rolle conditions check karo.
g , [ a , b ] par continuous hai (continuous f aur ek line ka sum). ✔
g , ( a , b ) par differentiable hai. ✔
Endpoints: g ( a ) = f ( a ) − f ( a ) = 0 aur g ( b ) = f ( b ) − [ f ( a ) + b − a f ( b ) − f ( a ) ( b − a ) ] = f ( b ) − f ( b ) = 0 . Isliye g ( a ) = g ( b ) = 0 . ✔
Ye step kyon? Secant subtract karne se endpoints same value par flatten ho jaate hain, jo Rolle ki hypothesis 3 hai.
Step 3 — g par Rolle apply karo. Ek aisa c ∈ ( a , b ) exist karta hai jahan g ′ ( c ) = 0 . Differentiate karo:
g ′ ( x ) = f ′ ( x ) − b − a f ( b ) − f ( a ) .
g ′ ( c ) = 0 set karo:
f ′ ( c ) = b − a f ( b ) − f ( a ) . ■
Ye step kyon? Rolle hume gap function ka ek flat point deta hai; flat gap ka matlab hai curve-slope = secant-slope.
Mnemonic MVT proof yaad karo
"Secant Subtract karo, Rolle ko Bulao."
Rolle bas MVT ka horizontal secant wala case hai (f ( a ) = f ( b ) ⇒ slope = 0 ).
Worked example Example 1 — MVT ke liye
c dhundho
f ( x ) = x 2 on [ 1 , 3 ] . c find karo.
Step 1: Secant slope = 3 − 1 f ( 3 ) − f ( 1 ) = 2 9 − 1 = 4.
Kyon? Ye woh target value hai jo f ′ ( c ) ko hit karni chahiye.
Step 2: f ′ ( x ) = 2 x . Set karo 2 c = 4 ⇒ c = 2.
Step 3: Check karo c = 2 ∈ ( 1 , 3 ) . ✔ x = 2 par tangent secant ke parallel hai.
Worked example Example 2 — Rolle in action
f ( x ) = x 2 − 4 x + 3 = ( x − 1 ) ( x − 3 ) on [ 1 , 3 ] .
Step 1: f ( 1 ) = 0 , f ( 3 ) = 0 , isliye f ( a ) = f ( b ) . ✔ Continuous & differentiable (polynomial). ✔
Kyon? Saari Rolle hypotheses hold hoti hain ⇒ andar flat tangent guaranteed hai.
Step 2: f ′ ( x ) = 2 x − 4 = 0 ⇒ c = 2 ∈ ( 1 , 3 ) . ✔
x = 2 par vertex flat point hai.
Worked example Example 3 — Application: function ko bound karna
Dikhao ki saare a , b ke liye ∣ sin b − sin a ∣ ≤ ∣ b − a ∣ .
Step 1: f ( x ) = sin x par MVT apply karo: kisi c ke liye b − a sin b − sin a = cos c .
Step 2: Absolute values lo: ∣ b − a ∣ ∣ sin b − sin a ∣ = ∣ cos c ∣ ≤ 1.
Kyon? Cosine 1 se bound hai, aur MVT difference ko ek single derivative value mein convert karta hai.
Step 3: Rearrange karo: ∣ sin b − sin a ∣ ≤ ∣ b − a ∣. ■
(Ye prove karta hai ki sin Lipschitz hai — real analysis ka ek key fact.)
Common mistake "Hypotheses optional hain — main waise bhi
c dhundh lunga."
Kyon sahi lagta hai: Zyaadatar textbook functions nice hote hain, isliye tum check karna bhool jaate ho.
Counterexample: f ( x ) = ∣ x ∣ on [ − 1 , 1 ] mein f ( − 1 ) = f ( 1 ) = 1 hai, lekin f ′ ( x ) = ± 1 kabhi 0 nahi hota. Rolle fail hota hai kyunki f x = 0 par differentiable nahi hai.
Fix: HAMESHA pehle [ a , b ] par continuity AUR ( a , b ) par differentiability verify karo.
Common mistake "MVT exactly ek
c guarantee karta hai."
Kyon sahi lagta hai: Simple examples mein ek unique c milta hai.
Sach: Ye kam se kam ek c guarantee karta hai; kai ho sakte hain. Jaise ek wavy curve ke kaafi saare tangents secant ke parallel ho sakte hain.
Fix: Ise "ek c exist karta hai " padho, na ki "ek unique c hai."
c ek endpoint ho sakta hai."
Kyon sahi lagta hai: Endpoints [ a , b ] ka hissa hain.
Sach: c open interval ( a , b ) mein rehta hai — strictly andar. Differentiability sirf andar chahiye (aur conclusion bhi sirf wahan claimed hai).
Fix: Hamesha c ko strictly a aur b ke beech report karo.
Recall Feynman: ek 12-saal ke bacche ko explain karo
Socho tum ek pahaad par chadhte ho aur wapas neeche uss dost ke paas aate ho jo usi height par khada hai jahan se tum chale the. Apni walk mein kahin na kahin, tum ek pal ke liye flat move kar rahe the — na upar, na neeche — pahaad ke bilkul top par. Ye Rolle's theorem hai.
Ab socho tum ek aisi jagah chalte ho jahan tumhara dost tumse zyaada upar khada hai. Tumhari "average steepness" koi number hai. MVT kehta hai ki ek pal par tumhare pair exactly usi average steepness par tilt the. Ise prove karne ke liye, hum cleverly "poora pahaad tilt karte hain" taaki end, start ki height se match kare, aur harder problem ko Rolle wale flat-hill version mein convert kar dete hain!
Rolle's Theorem ki 3 hypotheses kya hain? (1) f continuous on [ a , b ] ; (2) f differentiable on ( a , b ) ; (3) f ( a ) = f ( b ) .
Rolle's Theorem ka conclusion kya hai? There exists c ∈ ( a , b ) with f ′ ( c ) = 0 .
MVT ka conclusion batao. There exists c ∈ ( a , b ) with f ′ ( c ) = b − a f ( b ) − f ( a ) .
MVT ka geometric matlab kya hai? Koi tangent line endpoints ko join karne wali secant ke parallel hoti hai.
Rolle, MVT ka special case kaise hai? Jab f ( a ) = f ( b ) , secant slope 0 hoti hai, isliye f ′ ( c ) = 0 .
MVT ko Rolle se prove karne wali helper function kya hai? g ( x ) = f ( x ) − [ f ( a ) + b − a f ( b ) − f ( a ) ( x − a ) ] (curve minus secant).
MVT proof mein g ( a ) = g ( b ) = 0 kyon hai? Kyunki g secant se vertical gap measure karta hai, jo dono endpoints par zero hota hai.
Rolle ke proof ke liye max/min exist hone ki guarantee kaun sa theorem deta hai? The Extreme Value Theorem (continuity on a closed bounded interval).
Woh counterexample jahan non-differentiability ki wajah se Rolle fail hota hai? f ( x ) = ∣ x ∣ on [ − 1 , 1 ] : f ( − 1 ) = f ( 1 ) lekin f ′ kabhi 0 nahi hota.
Kya MVT ek unique c guarantee karta hai? Nahi — kam se kam ek c ; kai ho sakte hain.
MVT use karke ∣ sin b − sin a ∣ ko bound karo. ∣ sin b − sin a ∣ = ∣ cos c ∣ ∣ b − a ∣ ≤ ∣ b − a ∣ .
Extreme Value Theorem — Rolle ke proof ke liye max/min provide karta hai.
Fermat's Theorem (interior extrema) — interior extremum ⇒ f ′ = 0 , Rolle ka engine.
Continuity aur Differentiability — required hypotheses.
Increasing and Decreasing Functions — MVT use karke prove hota hai (f ′ > 0 ⇒ increasing).
Taylor's Theorem — MVT ko higher derivatives se generalize karta hai.
Cauchy's Mean Value Theorem — do functions ke liye MVT; L'Hôpital's rule tak le jaata hai.
Lipschitz Continuity — bounded derivative ⇒ Lipschitz, MVT se.
Fermat's Theorem: interior extremum
Continuous on closed interval
Differentiable on open interval
Secant slope over interval
Tangent parallel to secant