Inverse of 2×2 matrix: Visual walkthrough — Hinglish deep dive

2.6.10 · D2 · Maths › Matrices & Determinants — Introduction › Inverse of 2×2 matrix

Step 0 — Woh words jinhe hum use kar sakte hain

Symbols aane se pehle, har cheez ki pictures par agree kar lete hain.

Picture mein: blue arrow tumhara original $x$ hai. Machine $A$ usse stretch aur tilt karke orange arrow $A x$ banati hai. Green dashed arrow $A^{- 1}$ hai jo orange ko blue par wapas le jaata hai. Agar undo perfectly kaam kare, toh $A$ lagaane ke baad $A^{- 1}$ lagaane par tum exactly wahan pahuncho jahan se shuru kiya tha — yahi matlab hai $A^{- 1} A = I$ ka.

Step 1 — Goal ko ek picture-equation ke roop mein likho

KYA. Hum chahte hain ek machine $A^{- 1}$ taaki $A$ karne ke baad $A^{- 1}$ karne se kuch bhi change na ho.

KYUN. "Inverse" ka koi matlab nahi jab tak hum ise pin down na karein. Iska ek honest definition yeh hai: woh machine jo $A$ ke saath compose hokar do-nothing machine de. Isliye hum demand karte hain

A A^{- 1} = I .

PICTURE. Neeche, ek row mein teen grids: starting grid, grid jab $A$ ne use warp kiya, aur grid jab $A^{- 1}$ ne use wapas un-warp kiya start par. Sabse left aur sabse right wali grids identical hain — yahi identity hai jo equation $A A^{- 1} = I$ claim kar rahi hai.

Yahan har symbol apni jagah earn karta hai:

$A$ — known machine, $(a c b d)$ , chaar diye hue numbers.
$A^{- 1} = (p r q s)$ — unknown machine; $p, q, r, s$ woh chaar numbers hain jo humein find karne hain.
$I = (1001)$ — do-nothing target.

Step 2 — Ek matrix equation ko chaar number equations mein todna

KYA. Dono machines ko multiply karo (dekho Matrix multiplication) aur entry by entry match karo.

KYUN. Ek matrix equation secretly chaar ordinary equations hoti hain, ek per slot. Numbers ko matrices se solve karna zyada easy hai, isliye hum ise todh dete hain.

(a c b d) (p r q s) = (1001)

Har output slot padhte hue ( $A$ ki row times $A^{- 1}$ ka column):

(1) top-left a p + b r = 1 (2) top-right a q + b s = 0

(3) bottom-left c p + d r = 0 (4) bottom-right c q + d s = 1

PICTURE. Figure shade karta hai ki $A$ ki kaunsi row $A^{- 1}$ ke kaunse column se milti hai taaki chaar target slots mein se ek produce ho — isliye tum samajh sakte ho kyun equation (1) mein $q$ ya $s$ nahi hai (woh doosre column mein rehte hain).

Split notice karo: equations $(1), (3)$ mein sirf $p$ aur $r$ hain; equations $(2), (4)$ mein sirf $q$ aur $s$ hain. Do independent chhote puzzles.

Step 3 — Pehla chhota puzzle solve karo $p$ aur $r$ ke liye

KYA. $(1)$ aur $(3)$ ko saath use karo.

KYUN. Unmein exactly do unknowns $p, r$ hain — ek clean 2-equation, 2-unknown system.

$(3)$ se: $c p + d r = 0$ . $r$ solve karo:

r = - \frac{c}{d} p (abhi ke liye assume karo d \neq = 0)

Yahan symbol $r$ ko $p$ ke terms mein express kiya ja raha hai, isliye equation $(1)$ single-unknown ban jaati hai. $(1)$ mein substitute karo:

a p + b (- \frac{c}{d} p) = 1 ⟹ p = \frac{a d - b c}{d} (a - \frac{b c}{d}) = 1.

Bracket collapse hokar $\frac{a d - b c}{d}$ ban jaata hai. Toh

p = \frac{d}{a d - b c}, r = - \frac{c}{a d - b c} .

PICTURE. Figure substitution ko ek "funnel" ki tarah dikhata hai: do equations ek mein pour hoti hain, messy fraction $a - \frac{b c}{d}$ simplify hoti hai, aur $p$ bahar aata hai. Quantity $a d - b c$ pehli baar appear hoti hai — highlighted, kyunki yeh aage har line mein aayegi.

$a d - b c$ — baar baar aane wala combination. Ise Step 5 mein naam denge.
$r$ mein minus seedha equation $(3)$ ko rearrange karne ke minus se aaya — final formula mein jo sign-flip tum dekhoge uska yahi origin hai.

Step 4 — Doosra chhota puzzle solve karo $q$ aur $s$ ke liye

KYA. Step 3 repeat karo $(2)$ aur $(4)$ ke saath.

KYUN. Identical structure hai, isliye identical method — koi nayi ideas nahi chahiye, bas relabelled letters.

$(2)$ se: $a q + b s = 0 \Rightarrow q = - \frac{b}{a} s$ . $(4)$ mein substitute karo:

c (- \frac{b}{a} s) + d s = 1 ⟹ s \cdot \frac{a d - b c}{a} = 1 ⟹ s = \frac{a}{a d - b c}, q = - \frac{b}{a d - b c} .

PICTURE. Step 4 ke funnel ka side-by-side mirror, taaki tum literally symmetry dekh sako: $p \to s$ mein $d \to a$ swap hota hai, aur $r \to q$ mein $c \to b$ swap hota hai. Diagonal ka swap theek yahan paida ho raha hai.

Chaar saare answers collect karo:

p = \frac{d}{a d - b c}, q = \frac{- b}{a d - b c}, r = \frac{- c}{a d - b c}, s = \frac{a}{a d - b c} .

Step 5 — Answer package karo aur divisor ko naam do

KYA. Shared $\frac{1}{a d - b c}$ ko aage nikalo.

KYUN. Chaaon entries ek hi cheez se divide hoti hain. Isse factor out karne par pattern — swap, negate — clearly dikhta hai.

A^{- 1} = \frac{1}{a d - b c} (d - c - b a)

PICTURE. Figure original $A$ aur uske inverse ko overlay karta hai: ek curved blue arrow dono diagonal entries swap karta hai; do red minus-badges $b$ aur $c$ par lagte hain; ek green ring shared divisor label karta hai. Matrix $(d - c - b a)$ adjugate hai (dekho Adjugate and cofactors).

Step 6 — Degenerate case: agar $det A = 0$ ho toh?

KYA. Poocho ki kya hota hai jab $a d - b c = 0$ ho.

KYUN. Humne $a d - b c$ se chaar baar divide kiya. Agar yeh zero ho, toh har entry $\frac{number}{0}$ ban jaati hai — undefined. Undo machine exist nahi kar sakti. Humein geometrically dikhana hoga kyun, sirf algebraically nahi.

PICTURE. Jab $a d - b c = 0$ hota hai, toh $A$ ke dono columns ek hi line ki taraf point karte hain. Machine $A$ poori 2D grid ko us single line par squash kar deti hai: area zero ho jaata hai. Do alag input arrows ek hi output arrow par ja sakte hain — toh "undo" yeh nahi jaanta ki kisko wapas bhejna hai. Information hamesha ke liye lost ho jaati hai. Yeh ek singular matrix hai (dekho Singular vs non-singular matrices aur Linear transformations and area).

Step 7 — Confirm karo ki machine sach mein undo karti hai (chaar slots)

KYA. $A$ ko hamare $A^{- 1}$ se multiply karo aur check karo ki hume har slot mein $I$ milta hai.

KYUN. Humne answer forecast kiya; ab verify karte hain — sign slip ke khilaf sasta insurance (rule: har case cover karo, har entry including).

A A^{- 1} = \frac{1}{a d - b c} (a c b d) (d - c - b a) = \frac{1}{a d - b c} (a d - b c c d - d c - ab + ba - c b + d a)

= \frac{1}{a d - b c} (a d - b c 0 0 a d - b c) = (1001) = I ✓

Term by term: off-diagonal slots zero ho jaate hain kyunki $- ab + ba = 0$ aur $c d - d c = 0$ ; diagonal slots dono $a d - b c$ ban jaate hain, jise front factor cancel karke $1$ bana deta hai. Yeh exactly do-nothing machine hai — undo kaam karta hai.

Ek-picture summary

Sab kuch ek canvas par: goal $A A^{- 1} = I$ → chaar equations → solve karo → swap/negate/divide → $det A = 0$ par degenerate wall.

Recall Feynman retelling — poora walkthrough simple words mein

Main ek aisi machine chahta tha jo $A$ jo karta hai use undo kare. "Undo" ka matlab tab hi kuch hota hai jab dono milake kuch change na karein, toh maine demand kiya $A A^{- 1} = I$ aur unknown machine ko chaar mystery numbers $p, q, r, s$ se likha. Dono machines multiply karne par chaar chhote number-puzzles mile. Unme se do sirf $p$ aur $r$ ke baare mein jaante the, baaki do sirf $q$ aur $s$ ke baare mein — toh maine har pair apne aap solve ki. Har answer kuch original entry ko usi magic number $a d - b c$ se divide karke aaya. Woh magic number aage nikaalte hi leftovers ne ek neat rule spell kiya: do diagonal numbers swap karo, baaki do par minus signs lagao. Woh leftover grid adjugate hai; magic number determinant hai. Last sawaal: agar magic number zero ho toh? Tab main zero se divide kar raha hota — impossible. Aur yeh sense bhi banta hai: zero determinant ka matlab hai $A$ ne poore plane ko ek line par flatten kar diya, toh do alag arrows ek hi jagah end up ho sakte hain, aur koi undo machine unhe kabhi alag nahi bata sakti. Zero determinant, no undo. Baaki sab ke liye, boxed recipe kaam karta hai.

Active recall

A^{- 1}

derive karne ke liye hum kaunsi defining demand impose karte hain?

A

karne ke baad

A^{- 1}

karne se kuch change nahi hota, yaani

A A^{- 1} = I

Matrix equation

A A^{- 1} = I

chaar number equations mein kyun split hoti hai?

Kyunki har output slot

A

ki ek row times

A^{- 1}

ka ek column ke barabar hoti hai, jo chaar independent scalar equations deta hai.

Kaunsi do equations

p

aur

r

solve karti hain, aur kyun?

Equations (1) aur (3), kyunki unme sirf

p

aur

r

hain (ek clean 2×2 system).

Final formula mein

b, c

par sign-flip kahan se aata hai?

Equations (2) aur (3) ko rearrange karne par jo minus signs aate hain unse, jaise

r = - \frac{c}{d} p

Formula

a d - b c

se divide kyun kar sakta hai jabki Step 3 ne

d \neq = 0

assume kiya tha?

Assumption sirf ek algebra path choose karna tha; asli obstruction

det A = 0

hai, aur koi bhi nonzero det same formula tak valid path admit karta hai.

Geometrically,

det A = 0

hone par inverse kyun nahi hota?

Columns parallel hote hain, toh

A

plane ko ek line par squash kar deta hai; alag inputs same output par map ho jaate hain aur information lost ho jaati hai.

Verification

A A^{- 1}

mein off-diagonal entries kyun vanish ho jaate hain?

Kyunki

- ab + ba = 0

aur

c d - d c = 0

Connections

Inverse of 2×2 matrix — woh parent result jise yeh page visually derive karta hai.
Matrix multiplication — $A A^{- 1} = I$ ko chaar equations mein split karne aur verify karne ke liye use hua.
Identity matrix — derivation ka do-nothing target.
Determinant of a 2×2 matrix — divisor $a d - b c$ jo Steps 3–4 mein paida hua.
Adjugate and cofactors — swap/negate grid jo Step 5 mein package hua.
Singular vs non-singular matrices — Step 6 mein $det A = 0$ wall.
Linear transformations and area — kyun zero determinant ka matlab crushed area hai.
Solving linear systems — payoff: $x = A^{- 1} b$ .