Direction fields and Euler's method — visual - numerical intuition first

4.6.2 · Maths › Ordinary Differential Equations

1. Direction field kya hota hai?

WHY it works: ODE ek constraint hai — "yahan teri slope $f (x, y)$ ke barabar honi chahiye." Segment us constraint ko sirf visualise karta hai. Solution koi bhi aisi curve hai jo har jagah constraint obey kare.

2. Euler's method — scratch se derive karo

Hum ODE ko formulas se hamesha solve nahi kar sakte. Lekin hum arrows ko numerically follow kar sakte hain.

Derivation (first principles — local linear approximation):

Derivative hi tangent line ki slope hai. $x_{n}$ ke kareeb, true solution apni tangent se achhi tarah approximate hoti hai: $y (x_{n} + h) \approx y (x_{n}) + h \cdot y^{'} (x_{n}) .$

Yeh sirf first-order Taylor expansion hai: $y (x_{n} + h) = y (x_{n}) + h y^{'} (x_{n}) + \frac{h ^{2}}{2} y^{''} (ξ) .$

Lekin ODE hamein $y^{'} (x_{n}) = f (x_{n}, y_{n})$ deta hai. Substitute karo aur $h^{2}$ remainder ko drop karo:

3. Worked example: haath se chalao

Figure — Direction fields and Euler's method — visual - numerical intuition first

4. Common mistakes (Steel-man + fix)

5. Active recall

Recall Quick self-test (answers hide karo)

Q: $y^{'} = f (x, y)$ mein $f (x, y)$ physically kya deta hai? → $(x, y)$ par solution ki slope.
Q: Euler update formula? → $y_{n + 1} = y_{n} + h f (x_{n}, y_{n})$ .
Q: Global error ka order? → $O (h)$ (first order).
Q: Euler slope kahan evaluate karta hai? → current/left point par.

Recall Feynman: 12-saal ke bacche ko samjhao

Socho tum fog mein chal rahe ho aur ek magic compass tumhe batata hai ki "neeche ki taraf" exactly kidhar hai jahan tum khade ho. Tum poori pahadi nahi dekh sakte, lekin tum us direction mein ek chota step lete ho, compass se phir poochte ho, ek aur chota step lete ho, aur aage bhi yahi karte rehte ho. Jo path tum trace karte ho woh trail ki tumhari best guess hai. ODE magic compass hai; Euler's method chote steps lena hai. Chote steps = asli trail ke kareeb, lekin zyada kaam.

6. Connections

Separable First-Order ODEs — Euler se compare karne ke liye exact solutions.
Taylor's Theorem — $h y^{'}$ approximation aur $O (h^{2})$ remainder ka source.
Runge-Kutta Methods (RK4) — higher-order successors jo har step mein multiple slopes sample karte hain.
Existence and Uniqueness (Picard–Lindelöf) — jab ek point se ek single solution curve exist karti hai.
Stability and Stiff Equations — jab explicit Euler explode karta hai aur implicit methods chahiye hote hain.

What does

f (x, y)

y^{'} = f (x, y)

tell you geometrically?

(x, y)

point par solution curve ki slope.

Define a direction (slope) field.

Har point

(x, y)

par slope

f (x, y)

ka ek chota segment; solutions woh curves hain jo in segments ke har jagah tangent hain.

What is an isocline?

f (x, y) = m

ka locus jahan sab field segments ek hi slope

m

share karte hain.

State Euler's method update.

y_{n + 1} = y_{n} + h f (x_{n}, y_{n})

with

x_{n + 1} = x_{n} + h

Derive Euler from Taylor.

y (x_{n} + h) = y_{n} + h y_{n}^{'} + \frac{h ^{2}}{2} y^{''} (ξ)

;

y_{n}^{'} = f (x_{n}, y_{n})

replace karo aur

O (h^{2})

term drop karo.

What is the local truncation error of Euler?

\frac{h ^{2}}{2} y^{''} (ξ)

, yaani

O (h^{2})

per step.

What is the global error order of Euler?

O (h)

— first order;

h

half karne se total error roughly half hoti hai.

Which point's slope does explicit Euler use?

Current (left) point

(x_{n}, y_{n})

, naye point ki nahi.

Why does Euler undershoot a convex (

y^{''} > 0

) solution?

Tangent lines ek convex curve ke neeche rehti hain, to har step thoda neeche land karta hai.

Common error: omitting which factor?

Slope ko step

h

se multiply karna;

Δ y = h f

, sirf

f

nahi.