Scalar multiplication: Visual walkthrough — Hinglish deep dive

2.6.4 · D2 · Maths › Matrices & Determinants — Introduction › Scalar multiplication

Hum pehle kuch symbols build karenge, taaki kuch bhi tumhe surprise na kare:

Step 1 — Ek number aur ek arrow se shuru karo

KYA. Ek second ke liye matrices bhool jao. Ek single number $a$ lo. Ise $k$ se scale karna matlab hai product $k \cdot a$ — yeh toh sirf ordinary multiplication hai.

KYU. Ek matrix bana hota hai single numbers se. Agar hum agree nahi kar sakte ki "ek number scale karna" ka kya matlab hai, toh hum poora grid scale nahi kar sakte. Toh hum pehle simplest case pin down karte hain.

PICTURE. Number $a$ ko ek line par $a$ length ke arrow ki tarah draw karo. $k = 2$ se multiply karna arrow ko do guna lamba bana deta hai. $k = \frac{1}{2}$ se multiply karna use aadha kar deta hai. Woh stretch hi sab kuch hai jo scaling ek single value ke saath karti hai.

the stretch factor k \cdot the original number a = stretched number k a

Step 2 — Ek negative number ya zero ke saath kya hota hai?

KYA. Hume idea par trust karne se pehle awkward inputs check karne chahiye. Agar $a$ negative hai, toh uska arrow left ki taraf point karta hai. Agar $k$ negative hai, toh woh arrow ko $0$ ke doosri taraf flip karta hai aur phir stretch karta hai. Agar $k = 0$ hai, toh arrow origin par ek single point mein collapse ho jaata hai.

KYU. Jo rule tumne sirf achhe positive numbers par test kiya hai woh rule hai jo pehli baar real data mein minus sign aane par toot jaayega. Hum abhi har sign cover kar lete hain taaki baad mein koi scenario surprise na kare.

PICTURE. Origin se teen arrows: $k = 2$ (lamba, same side), $k = - 1$ (same length, flipped side — ek reflection), $k = 0$ (kuch nahi tak shrink ho gaya).

Yahan $∣ k ∣$ (padho "size of $k$ ") sirf $k$ hai jisme se koi bhi minus sign hata diya gaya ho — yeh length measure karta hai, jo kabhi negative nahi hoti.

Step 3 — Numbers ko ek grid mein rakh do

KYA. Ab poori matrix ek saath dekhne ke liye char numbers ko $2 \times 2$ grid mein rakho:

A = [a_{11} a_{21} a_{12} a_{22}]

KYU. Matrix hamaara container hai. Isse scale karne se pehle, hume bilkul clear hona chahiye ki har entry ek fixed cell mein rehti hai, aur har cell Step 1 ke numbers mein se sirf ek hai.

PICTURE. Char cells ka ek grid, har ek apne address $a_{ij}$ se labelled. Accent colour ek cell, $a_{21}$ (row 2, column 1), mark karta hai, taaki tum agle steps mein usi exact cell ko dekh sako.

one number a_{ij} \in the grid holding all of them A

Step 4 — Ek cell scale karo, baaki ignore karo

KYA. Sirf highlighted cell $a_{21}$ lo aur Step 1 use karke use $k$ se scale karo. Woh $k \cdot a_{21}$ ho jaata hai. Abhi kuch aur nahi hilta.

KYU. Yahan key insight hai jo parent note ne state ki thi par show nahi ki thi: cells aapas mein baat nahi karte. Ek cell ki value ka doosre cell ke scale hone par koi effect nahi hota. Toh grid ko scale karna aur kuch nahi balki single-number rule ko cell by cell repeat karna hai.

PICTURE. Wahi grid; red cell ka number $k a_{21}$ se replace ho jaata hai jabki teen black cells untouched baithe hain — grey, apni baari ka wait karte hue.

(k A)_{21} = stretch k \cdot this cell only a_{21}

Symbol $(k A)_{21}$ ka matlab hai "result $k A$ ki row 2, column 1 mein entry."

Step 5 — Ek saath har cell par karo

KYA. Step 4 ko charon cells ke liye simultaneously repeat karo. Har $a_{ij}$ apni jagah par $k a_{ij}$ ho jaata hai.

KYU. Kyunki cells independent hain (Step 4), "matrix scale karna" ka matlab sirf "har cell scale karna" ho sakta hai — scale karne ke liye aur kuch hai hi nahi. Yeh poora definition hai, ab hum par force kiya gaya rather than declare kiya gaya.

PICTURE. Charon cells ek saath red ho jaate hain, har ek $k a_{ij}$ dikhata hai. Grid ki shape — 2 rows, 2 columns — identical hai; sirf contents badla.

General statement (koi bhi size $m \times n$ ) exactly yahi idea hai jo zyada cells mein copy hua hai:

(k A)_{ij} = k \cdot a_{ij} for every i, j

Step 6 — Shape nahi badal sakta (ek degenerate check)

KYA. Poochho: kya scaling kabhi ek row add kar sakti hai, column delete kar sakti hai, ya cell move kar sakti hai? Nahi. $k$ se multiplication sirf har cell ke andar ka number badalta hai.

KYU. Yeh ek common fear khatam karta hai (parent mein Mistake 3 dekho): scaling ek contents operation hai, structure operation nahi. Ek $m \times n$ matrix $m \times n$ rehti hai. Yahan tak ki extreme case $k = 0$ shape rakhti hai — tumhe zeros ka poora $m \times n$ grid milta hai (zero matrix $O$ ), na ki ek empty matrix.

PICTURE. Left: $A$ apne outline ke saath. Right: $k A$ same outline ke saath par scaled numbers. Ek red bracket dikhata hai outline bilkul identical hai.

A is m \times n ⟹ k A is m \times n

$k = 1$ : har cell unchanged, toh $1 \cdot A = A$ .
$k = 0$ : har cell $0$ ho jaata hai, toh $0 \cdot A = O$ (same shape, sab zeros).

Step 7 — Ek real example ko transform hote dekho

KYA. Maano $A = [25 - 3 0]$ aur $k = 4$ se scale karo.

KYU. Yeh prove karne ke liye ki picture aur arithmetic agree karte hain, aur ek negative entry ( $- 3$ ) aur ek zero entry ( $0$ ) include karne ke liye taaki har tarah ki cell dikhe.

PICTURE. $A$ ki har cell ka apni length ka ek arrow hai; har arrow $4 \times$ apni length tak badh jaata hai, aur $- 3$ wala arrow (negative direction mein point karta hua) badh ta hai jabki apni direction rakhta hai. Zero cell ek dot rehti hai.

4 [25 - 3 0] = [4 \cdot 2 4 \cdot 5 4 \cdot (- 3) 4 \cdot 0] = [820 - 12 0]

Notice karo: $8$ badha, $- 12$ badha aur negative raha, $20$ badha, $0$ $0$ raha. Step 2 ka har case ek matrix mein appear karta hai.

Recall Khud check karo

Compute karo $\frac{1}{3} [612 - 9 15]$ . ::: $[24 - 3 5]$ — har cell $\frac{1}{3}$ se multiply hua; 1 se chota scalar shrink karta hai. $0 \cdot [7129]$ kya hai, aur uski shape kya hai? ::: Zero matrix $[0000]$ , abhi bhi $2 \times 2$ .

Ek-picture summary

Upar ki sab cheez ek single flow mein compress hoti hai: ek number scale hota hai (Step 1–2), grid aisi bahut saari numbers side by side hai (Step 3), aur kyunki cells kabhi interact nahi karte, grid ko scale karna hai har cell ko jagah par scale karna jabki bahari shape fixed rehti hai (Steps 4–6).

Yeh exactly ek linear transformation ka seed bhi hai (space ki uniform stretching), aur yeh addition ke saath properties of matrix operations mein collect kiye gaye distributive laws ke through cleanly kaam karta hai. Yeh matrix multiplication ka row-by-column combining nahi hai — woh ek alag, mushkil operation hai.

Recall Feynman retelling — poori walkthrough plain words mein

Socho ek tic-tac-toe board jisme har square mein ek number likha hai. Koi tumhe ek "stretch factor" deta hai, maano $4$ . Yahan woh trick hai jo main chahta hoon tum dekho, sirf believe na karo: har square ka number ek chota arrow hai, aur har square bilkul akela hai — top-left square ko bilkul nahi pata ki bottom-right square mein kya hai. Toh tum bas har square par jaate ho aur uske arrow ko $4$ se stretch kar dete ho. Positive numbers lamba hote hain. Negative numbers lamba hote hain par negative direction mein point karte rehte hain. Zero sirf ek dot rehta hai chahe kuch bhi ho. Jab tum khatam karte ho, board mein exactly utni hi rows aur columns hoti hain jitni pehle thi — tumne kabhi ek square add ya remove nahi kiya, tumne sirf andar jo tha use rewrite kiya. Agar tumhara stretch factor $1$ hota, kuch nahi badlta; agar $0$ hota, har square $0$ ho jaata par board abhi bhi wahan hota, sirf zeros se bhara hua. Yahi $k A$ ki poori kahaani hai: same board, har arrow $k$ se stretch hua.