4.3.13 · D3 · Maths › Calculus III — Sequences & Series › Root test
Yeh Root test ka child deep-dive hai. Yahan hum theorem re-derive nahi karte; hum har possible shape ko dhundh ke nikalte hain jo ek root-test problem le sakti hai aur har ek ke liye ek example solve karte hain. Agar exam mein koi series mile, toh woh neeche ki rows mein se kisi ek mein zaroor aayegi.
Poore note mein, a n ko series ∑ a n ka n -waan term maano — sum ke n -wein slot mein baithne wala number (∑ n 2 1 ke liye, term a n hai n 2 1 ). Har root-test problem ek single number se decide hoti hai:
L = lim n → ∞ n ∣ a n ∣ .
Yahan n x ("x ka n -waan root") woh number hai jo khud ko n baar multiply karne par x deta hai. Yeh power n ko undo karta hai. Yahi reason hai ki hum ise use karte hain: agar ek term secretly r n ke barabar hai, toh n r n = r chhupa hua ratio r wapas bahar le aata hai. Hum L compute karte hain, phir verdict padhte hain: L < 1 matlab converges, L > 1 matlab diverges, L = 1 kuch nahi bolta.
Neeche ke cells har tarah ka input list karte hain jo is machine ko mil sakta hai. Example label par click karke uski full working tak jump karo.
Cell
Situation
Kya cheez ise alag banati hai
Example
A
Clean ( stuff ) n , ratio < 1
root power ko cancel karta hai, L < 1
Ex 1
B
Exponential over polynomial, base > 1
L > 1 divergence
Ex 2
C
Polynomial over exponential
polynomial root hokar 1 banta hai, base < 1 bachta hai
Ex 3
D
L = 1 degenerate (borderline)
test chup hai, tools switch karne padenge
Ex 4
E
L = 0 super-fast decay
ratio effectively 0 , strongly converges
Ex 5
F
L = ∞ blow-up
terms explode ho jaate hain, diverges
Ex 6
G
Negative / alternating terms
$
a_n
H
Word problem (bouncing / physical)
reality ko series mein translate karo
Ex 8
I
Exam twist: n exponent ke exponent mein
( 1 + n c ) n → e c
Ex 9
Hum sab ko cover karenge.
Worked example Ex 1 (Cell A) ·
n = 1 ∑ ∞ ( 5 n + 2 4 n − 1 ) n
Forecast: poori bracket n ki power par raise ki gayi hai. Guess karo: andar ka fraction 1 se neeche, 1 par, ya 1 se upar settle hoga? Padhne se pehle apna guess likh lo.
Root apply karo. n ∣ a n ∣ = n ( 5 n + 2 4 n − 1 ) n = 5 n + 2 4 n − 1 .
Yeh step kyun? Term ek perfect n -th power hai, isliye n ( ⋅ ) n = ( ⋅ ) — root aur power ek doosre ko annihilate kar dete hain. Yeh bilkul woh situation hai jiske liye root test bana hai.
Limit lo. Top aur bottom dono ko n se divide karo: 5 + 2/ n 4 − 1/ n → 5 4 .
Yeh step kyun? Jab n badhta hai, 1/ n → 0 aur 2/ n → 0 ; sirf leading coefficients bachte hain.
Verdict padho. L = 5 4 < 1 ⇒ absolutely converges.
Yeh step kyun? L < 1 ka matlab hai terms eventually ek convergent geometric series ∑ r n se bhi tezi se shrink hoti hain jahan 5 4 < r < 1 .
Verify karo: n = 100 plug karo: 502 399 = 0.7948 ≈ 0.8 . Trend confirm karta hai L = 0.8 < 1 . ✓
Worked example Ex 2 (Cell B) ·
n = 1 ∑ ∞ n 4 5 n
Forecast: top 5 n jaisa badhta hai, bottom n 4 jaisa. Kaun jeeta hai? Convergence ya divergence guess karo.
Har factor ka root lo. n n 4 5 n = n n 4 n 5 n = ( n n ) 4 5 .
Yeh step kyun? n 5 n = 5 exactly hai, aur n n 4 = ( n 1/ n ) 4 — hum base aur polynomial ko alag alag root karte hain.
Standard limit use karo n n → 1 : toh ( n n ) 4 → 1 4 = 1 , aur L = 1 5 = 5 milta hai.
Yeh step kyun? n 1/ n = e l n n / n aur n l n n → 0 , isliye n 1/ n → e 0 = 1 . Polynomial factors hamesha root hokar 1 bante hain; sirf sach mein exponential base 5 bachta hai.
Verdict. L = 5 > 1 ⇒ diverges.
Verify karo: a n = 5 n / n 4 ; check karo terms badhti hain: a 5 = 3125/625 = 5 , a 10 = 5 10 /1 0 4 ≈ 976562 . Terms explode kar rahe hain, toh a n → 0 — divergence confirm. ✓
Worked example Ex 3 (Cell C) ·
n = 1 ∑ ∞ 4 n n 3 + 2 n
Forecast: exponential ab neeche hai. Outcome guess karo.
Root ko split karo. n 4 n n 3 + 2 n = 4 n n 3 + 2 n .
Yeh step kyun? n 4 n = 4 ; base 4 root se bina bade nikal jaata hai.
Polynomial ka root lo. Bade n ke liye, n 3 + 2 n ≈ n 3 , toh n n 3 + 2 n → n n 3 = ( n n ) 3 → 1.
Yeh step kyun? + 2 n , n 3 ke against negligible hai; aur koi bhi polynomial Ex 2 ki tarah root hokar 1 banta hai. Isliye numerator ka root → 1 .
Verdict. L = 4 1 < 1 ⇒ converges.
Verify karo: 4 1 = 0.25 < 1 . Sanity: a 10 = ( 1000 + 20 ) / 4 10 ≈ 9.7 × 1 0 − 4 , already tiny aur shrink ho raha hai. ✓
Worked example Ex 4 (Cell D) ·
n = 1 ∑ ∞ n ( ln n ) 2 1 (n = 2 se start)
Forecast: yahan koi obvious n -th power nahi hai. Guess karo ki root test yeh decide bhi kar payega ya nahi.
Root compute karo. n n ( ln n ) 2 1 = n 1/ n ( ln n ) 2/ n 1 .
Yeh step kyun? Har factor ka root lo: n n = n 1/ n aur n ( ln n ) 2 = ( ln n ) 2/ n .
Dono roots → 1 . n 1/ n → 1 ; aur ( ln n ) 2/ n = e n 2 l n ( l n n ) → e 0 = 1 kyunki n l n ( l n n ) → 0 .
Yeh step kyun? Jo bhi cheez n mein exponential se dheemar badhti hai, woh root hokar 1 banti hai. Dono factors qualify karte hain.
Verdict. L = 1 ⇒ inconclusive — root test decide nahi kar sakta.
Yeh step kyun? L = 1 blind spot hai: yeh convergent aur divergent tails mein fark nahi kar sakta.
Tools switch karo. Comparison test integral ∫ 2 ∞ x ( l n x ) 2 d x = [ − l n x 1 ] 2 ∞ = l n 2 1 (finite) ke saath dikhata hai ki series converges .
Verify karo: antiderivative check: d x d ( − l n x 1 ) = x ( l n x ) 2 1 . ✓ Aur L = 1 exactly boundary par hai. ✓
Worked example Ex 5 (Cell E) ·
n = 1 ∑ ∞ n n 1
Forecast: exponent khud n hai. Guess karo L kitna chhota hoga.
Root lo. n n n 1 = ( n n ) 1/ n 1 = n 1 .
Yeh step kyun? ( n n ) 1/ n = n n / n = n 1 = n . Root andar ka honest n 1 expose karta hai.
Limit. n 1 → 0 , isliye L = 0 .
Yeh step kyun? Yahan effective geometric ratio sirf 1 se neeche nahi — yeh 0 par collapse ho jaata hai, isliye terms kisi bhi fixed r n se tezi se vanish hoti hain.
Verdict. L = 0 < 1 ⇒ converges (bahut strongly).
Verify karo: a 1 = 1 , a 2 = 4 1 , a 3 = 27 1 , a 4 = 256 1 — tezi se gir raha hai. Partial sum already ≈ 1.29 aur barely move kar raha hai. ✓
Worked example Ex 6 (Cell F) ·
n = 1 ∑ ∞ n n
Forecast: Ex 5 ka mirror image. L kya karta hai?
Root lo. n ∣ n n ∣ = ( n n ) 1/ n = n .
Yeh step kyun? Ex 5 jaisi hi algebra, lekin reciprocal ki jagah term khud ke saath.
Limit. n → ∞ , isliye L = ∞ .
Verdict. L = ∞ > 1 ⇒ diverges.
Yeh step kyun? n-th term divergence test already chilla raha hai: n n → ∞ = 0 , toh convergence ka koi chance nahi.
Verify karo: a 1 = 1 , a 2 = 4 , a 3 = 27 , a 4 = 256 — explode ho raha hai, a n → 0 . ✓
Worked example Ex 7 (Cell G) ·
n = 1 ∑ ∞ ( − 1 ) n ( 2 n + 1 n ) n
Forecast: sign har term flip hoti hai. Kya root test ko fark padta hai?
Pehle absolute value lo. Root test ∣ a n ∣ use karta hai, aur ∣ ( − 1 ) n ∣ = 1 , isliye ∣ a n ∣ = ( 2 n + 1 n ) n .
Yeh step kyun? n ∣ a n ∣ sirf magnitudes dekhta hai — alternating sign mit jaati hai. Yahi reason hai ki root test directly absolute convergence prove karta hai.
Root aur limit. n ∣ a n ∣ = 2 n + 1 n = 2 + 1/ n 1 → 2 1 .
Yeh step kyun? Perfect n -th power phir se; andar n se divide karo.
Verdict. L = 2 1 < 1 ⇒ absolutely converges (aur converges bhi).
Verify karo: n = 99 : 199 99 = 0.4975 ≈ 0.5 . ✓ Sign ne kabhi root ko affect nahi kiya. ✓
Worked example Ex 8 (Cell H) · Ek bouncing signal
Ek radar pulse har hop mein power khoti hai. Hop n par woh E n = ( 4 3 ) n n 2 joules energy carry karta hai. Kya sabhi hops par total energy ek finite amount mein converge karti hai?
Forecast: 4 3 ise shrink karta hai, lekin n 2 push back karta hai. Kaun jeeta hai?
Series ke roop mein model karo. Total energy = ∑ n = 1 ∞ E n = ∑ n 2 ( 4 3 ) n .
Yeh step kyun? "Sabhi hops par total" literally har hop ki energy ka infinite sum hai.
Har factor ka root lo. n n 2 ( 4 3 ) n = ( n n ) 2 ⋅ 4 3 → 1 ⋅ 4 3 = 4 3 .
Yeh step kyun? Polynomial n 2 root hokar 1 banta hai (Ex 2 wali logic); geometric base 4 3 bachta hai.
Verdict. L = 4 3 < 1 ⇒ converges : total energy finite hai. Physical geometric decay 4 3 < 1 polynomial growth ko haara deta hai.
Verify karo (units + numerics): E n joules mein; joules ka sum joules hota hai — dimensionally theek hai. Terms: E 1 = 0.75 , E 10 ≈ 5.6 , E 30 ≈ 0.16 , E 50 ≈ 0.001 — pehle peak phir 0 ki taraf decay, total finite. ✓
Neeche ka figure yeh competition visible banata hai: cyan bars per-hop energies E n hain (woh ek early peak tak uthte hain, phir khatam ho jaate hain), jabki amber curve running total hai — dekho yeh ek finite plateau par flat hota hai, jo exactly wahi hai jo L = 4 3 < 1 guarantee karta hai.
Worked example Ex 9 (Cell I) ·
n = 1 ∑ ∞ ( 1 + n 3 ) n 2 1
Forecast: outer power n 2 hai, n nahi. Yeh classic trap hai. Guess karo.
Root lo — n 2 ban jaata hai n . n ( 1 + n 3 ) n 2 1 = ( 1 + n 3 ) n 2 / n 1 = ( 1 + n 3 ) n 1 .
Yeh step kyun? Root exponent ko n se divide karta hai: n 2 / n = n . Yeh famous limit shape ( 1 + n c ) n chhor jaata hai.
( 1 + n c ) n → e c use karo jahan c = 3 hai: denominator → e 3 banta hai, isliye L = e 3 1 = e − 3 .
Yeh step kyun? Yeh exponential e ki definition hai: ( 1 + n c ) n → e c . Hum e yahan isliye invoke karte hain kyunki ( 1 + small ) ko badi power par raise karna exactly wahi hai jo e measure karta hai.
Verdict. L = e − 3 ≈ 0.0498 < 1 ⇒ converges.
Verify karo: e − 3 ≈ 0.0498 < 1 . n = 1000 par numeric check: ( 1 + 3/1000 ) 1000 ≈ e 3 ≈ 20.09 , root ≈ 1/20.09 ≈ 0.0498 . ✓
Recall Matrix par khud ko test karo
Kaun sa cell: ∑ ( n / ( 3 n + 1 ) ) n ? ::: Cell A — clean power, L = 1/3 < 1 , converges.
Kaun sa cell: ∑ 7 n / n 10 ? ::: Cell B — exponential over polynomial, L = 7 > 1 , diverges.
Kaun sa cell: ∑ 1/ n p (koi bhi p )? ::: Cell D — L = 1 , inconclusive; p-series use karo.
Kaun sa cell: ∑ 1/ ( 2 n ) n ? ::: Cell E — L = 0 , strongly converges.
Kya alternating sign L change karta hai? ::: Nahi — root test ∣ a n ∣ use karta hai, isliye sign mit jaati hai (Cell G).
Mnemonic Poore matrix ka ek-line recap
Term ka root lo, polynomials ko 1 karo, asli base ko r rakho, aur L ko 1 se compare karo. n ke exponent mein powers divide out ho jaati hain; ( 1 + c / n ) n ban jaata hai e c .
Root test — parent theorem jise yeh page drill karta hai.
Geometric series — r n yardstick jisse har L compare hota hai.
Ratio test — Ex 9 par try karo aur takleef mehsoos karo; n -th powers par root jeetta hai.
Comparison test — L = 1 Cell D mein rescue tool.
n-th term divergence test — Cells B aur F mein fast kill.
p-series — canonical L = 1 specimens.