Foundations — Eigenvalues and eigenvectors — characteristic polynomial

4.5.29 · D1 · Maths › Linear Algebra (Full) › Eigenvalues aur eigenvectors — characteristic polynomial

Parent note padhne se pehle, tumhare paas symbols ka ek chhota toolbox hona chahiye. Neeche har symbol aur idea diya gaya hai jo parent use karta hai, bilkul zero se bana hua, ek aisi tartib mein jahan har cheez sirf usse pehle wali cheez pe depend karti hai.

1. Arrow: vector $\mathbf{x}$ kya hota hai

Figure dekho: laal arrow origin (point $(0,0)$) se shuru hota hai aur uski tip point $(x_1,x_2)$ par jaake rukti hai. Do kaali dashed legs "daayein" wali amount $x_1$ aur "upar" wali amount $x_2$ dikhati hain.

Woh special arrow jisme saare zeros hain, $0=\left(\genfrac{}{}{0ex}{}{0}{0}\right)$, zero vector kehlata hai — iske paas koi length nahi hai aur yeh kahin point nahi karta. Isse yaad rakho: parent note ise eigenvector ke roop mein forbid karta hai, aur hum dekhenge kyun.

2. Machine: matrix $A$ kya hoti hai

$A$ ko vector $\mathbf{x}$ se multiply karne par ek naya vector $A\mathbf{x}$ milta hai, row-by-row compute hota hai: $A\mathbf{x}=\left(\begin{array}{cc}a& b\\ c& d\end{array}\right)\left(\begin{array}{c}{x}_1\\ x_2\end{array}\right)=\left(\begin{array}{c}a{x}_1+b{x}_2\\ c{x}_1+d{x}_2\end{array}\right).$

Figure mein, kaala arrow input $\mathbf{x}$ hai; laal arrow output $A\mathbf{x}$ hai. Dhyan do yeh alag direction mein point kar raha hai — machine ne ise ghuma diya. Eigenvector woh special input hai jahan kaala aur laal ek hi line par hote hain.

3. Stretch number: scalar $\lambda$ kya hota hai

Har tarah ka $\lambda$ kaisa dikhta hai?

• $\lambda =2$: arrow do guna lamba, same direction.
• $\lambda =\frac{1}{2}$: arrow aadha lamba, same direction.
• $\lambda =-1$: same length, flip hokar opposite direction mein point karta hai.
• $\lambda =0$: zero vector mein collapse ho jaata hai.

Figure mein ek kaala arrow $2$ se, $-1$ se, aur $0$ se scale kiya gaya hai — teeno copies origin se guzarne wali same dashed line par rehti hain. Us line par rehna hi "sirf stretch, kabhi ghuma nahi" ka visual meaning hai, jo $A\mathbf{x}=\lambda \mathbf{x}$ demand karta hai.

4. "Kuch nahi karne wali" machine: identity matrix $I$

Parent note $\lambda I$ mein achanak $I$ kyun use karta hai? Kyunki tum ek plain number ko matrix se subtract nahi kar sakte: expression "$A-\lambda$" meaningless hai — ek grid minus ek single number ka koi defined rule nahi hai. Lekin identity ko scale karna, $\lambda I$, ek genuine matrix hai: $\lambda I=\left(\begin{array}{cc}\lambda & 0\\ 0& \lambda \end{array}\right).$ Ab $A-\lambda I$ ek legal grid-minus-grid subtraction hai. Yahi ek wajah hai ki $I$ appear hota hai.

5. Volume dial: determinant $\det$

Topic ko determinant ke zero hone ki care kyun hai? Zero area ka matlab hai parallelogram ek line par squash ho gaya — machine ne 2D ko 1D mein crush kar diya. Jab aisa hota hai machine singular hoti hai: yeh information destroy karti hai aur ise undo nahi kiya ja sakta. Poori story ke liye Determinants dekho.

Left: ek normal matrix — uske do column arrows (kaale) ek laal parallelogram span karte hain jiska real area hai. Right: ek singular matrix — columns line up ho gaye hain, laal parallelogram ek segment mein flatten ho gaya hai, area $=0$.

6. Arrow ko kuch nahi mein bhejna: null space aur singular matrices

Parent ka Step 3 kehta hai: "$M\mathbf{x}=0$ ka ek nonzero solution hai iff $M$ singular hai." Yahan simple words mein kyun:

• Agar $M$ undo kiya ja sakta (invertible), toh undo-machine $M^{-1}$ ko $M\mathbf{x}=0$ ke dono sides par apply karne se $\mathbf{x}=0$ milta hai — sirf boring zero arrow solution hai.
• Toh ek nonzero solution tab hi exist kar sakta hai jab $M$ undo nahi kiya ja sakta — yaani $M$ singular hai — yaani $\det M=0$.

Yeh chain, "nonzero solution $\iff$ singular $\iff$ determinant zero," woh engine hai jo eigen-equation ko ek solvable polynomial mein badal deta hai. Aur details Null Space and Rank mein hain.

7. Treasure map: polynomial aur uske roots kya hote hain

Ek $2\times 2$ matrix ke liye, $\det (A-\lambda I)$ ek degree-2 polynomial nikalta hai (iska highest power ${\lambda }^2$ hai); $n\times n$ ke liye iska degree $n$ hota hai. Uske roots eigenvalues hain. Toh poori hunt ek school-algebra task tak reduce ho jaati hai: ek polynomial ke roots dhundho.

8. Diagonal sum: trace $\mathrm{tr}(A)$

Yeh matter karta hai kyunki yeh eigenvalues ke sum ke barabar hota hai — har answer par ek one-line check.

9. Jab arrows cooperate nahi karte: complex numbers

Topic ko inki zaroorat kyun hai: kuch real matrices (jaise $90^{\circ }$ rotation) har arrow ko ghuma deti hain, toh koi bhi real arrow kabhi apni line par nahi rehta. Tab characteristic polynomial, jaise ${\lambda }^2+1=0$, ke koi real roots nahi hote — lekin iske complex roots $\pm i$ hote hain. Complex $\lambda$ allow karne se guarantee hoti hai ki treasure map mein hamesha khazana milega.

Prerequisite map

Isse top-down padho: arrows aur machines eigen-equation ko feed karte hain; identity hume $A-\lambda I$ banane deti hai; determinant aur null space explain karte hain ki yeh singular kab hoti hai; polynomials, trace aur complex numbers algebra complete karte hain — aur sab kuch topic mein aa jaata hai.

Equipment checklist

Daayein side chhupa lo aur khud test karo. Tum parent note ke liye ready ho jab har line obvious lagey.