{"id":2778,"date":"2025-06-06T07:36:11","date_gmt":"2025-06-06T07:36:11","guid":{"rendered":"https:\/\/diznr.com\/?p=2778"},"modified":"2025-06-06T07:36:11","modified_gmt":"2025-06-06T07:36:11","slug":"computer-science-numerical-methods-finite-difference-method-difference-backward","status":"publish","type":"post","link":"https:\/\/www.reilsolar.com\/pdf\/computer-science-numerical-methods-finite-difference-method-difference-backward\/","title":{"rendered":"Computer Science\/Numerical Methods\/ Finite difference method( backward difference)"},"content":{"rendered":"

Computer Science\/Numerical Methods\/ Finite difference method( backward difference)<\/p>\n

[fvplayer id=”108″]<\/p>\n

Finite Difference Method: Backward Difference<\/strong><\/h3>\n

The Finite Difference Method (FDM)<\/strong> is a numerical technique used for approximating derivatives. The backward difference method<\/strong> is a commonly used approach when calculating derivatives at a given point using previous data points.<\/p>\n

\u00a0Definition of Backward Difference<\/strong><\/h3>\n

The backward difference<\/strong> of a function f(x)f(x)<\/span>f<\/span>(<\/span>x<\/span>)<\/span><\/span><\/span><\/span> at a point xix_i<\/span>x<\/span>i<\/span><\/span><\/span>\u200b<\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span> with step size hh<\/span>h<\/span><\/span><\/span><\/span> is given by:<\/p>\n

\u0394bf(xi)=f(xi)\u2212f(xi\u22121)\\Delta_b f(x_i) = f(x_i) – f(x_{i-1})<\/span>\u0394b<\/span><\/span><\/span>\u200b<\/span><\/span><\/span><\/span><\/span>f<\/span>(<\/span>x<\/span>i<\/span><\/span><\/span>\u200b<\/span><\/span><\/span><\/span><\/span>)<\/span>=<\/span><\/span>f<\/span>(<\/span>x<\/span>i<\/span><\/span><\/span>\u200b<\/span><\/span><\/span><\/span><\/span>)<\/span>\u2212<\/span><\/span>f<\/span>(<\/span>x<\/span>i<\/span>\u2212<\/span>1<\/span><\/span><\/span>\u200b<\/span><\/span><\/span><\/span><\/span>)<\/span><\/span><\/span><\/span><\/span><\/p>\n

For higher-order differences<\/strong>, we define:<\/p>\n

\u0394b2f(xi)=\u0394bf(xi)\u2212\u0394bf(xi\u22121)\\Delta^2_b f(x_i) = \\Delta_b f(x_i) – \\Delta_b f(x_{i-1})<\/span>\u0394b<\/span><\/span>2<\/span><\/span><\/span>\u200b<\/span><\/span><\/span><\/span><\/span>f<\/span>(<\/span>x<\/span>i<\/span><\/span><\/span>\u200b<\/span><\/span><\/span><\/span><\/span>)<\/span>=<\/span><\/span>\u0394b<\/span><\/span><\/span>\u200b<\/span><\/span><\/span><\/span><\/span>f<\/span>(<\/span>x<\/span>i<\/span><\/span><\/span>\u200b<\/span><\/span><\/span><\/span><\/span>)<\/span>\u2212<\/span><\/span>\u0394b<\/span><\/span><\/span>\u200b<\/span><\/span><\/span><\/span><\/span>f<\/span>(<\/span>x<\/span>i<\/span>\u2212<\/span>1<\/span><\/span><\/span>\u200b<\/span><\/span><\/span><\/span><\/span>)<\/span><\/span><\/span><\/span><\/span> \u0394b3f(xi)=\u0394b2f(xi)\u2212\u0394b2f(xi\u22121)\\Delta^3_b f(x_i) = \\Delta^2_b f(x_i) – \\Delta^2_b f(x_{i-1})<\/span>\u0394b<\/span><\/span>3<\/span><\/span><\/span>\u200b<\/span><\/span><\/span><\/span><\/span>f<\/span>(<\/span>x<\/span>i<\/span><\/span><\/span>\u200b<\/span><\/span><\/span><\/span><\/span>)<\/span>=<\/span><\/span>\u0394b<\/span><\/span>2<\/span><\/span><\/span>\u200b<\/span><\/span><\/span><\/span><\/span>f<\/span>(<\/span>x<\/span>i<\/span><\/span><\/span>\u200b<\/span><\/span><\/span><\/span><\/span>)<\/span>\u2212<\/span><\/span>\u0394b<\/span><\/span>2<\/span><\/span><\/span>\u200b<\/span><\/span><\/span><\/span><\/span>f<\/span>(<\/span>x<\/span>i<\/span>\u2212<\/span>1<\/span><\/span><\/span>\u200b<\/span><\/span><\/span><\/span><\/span>)<\/span><\/span><\/span><\/span><\/span><\/p>\n

\u00a0Backward Difference Approximation for Derivatives<\/strong><\/h3>\n

The first derivative using backward difference<\/strong> is approximated as:<\/p>\n

f\u2032(xi)\u2248f(xi)\u2212f(xi\u22121)hf'(x_i) \\approx \\frac{f(x_i) – f(x_{i-1})}{h}<\/span>f<\/span>\u2032<\/span><\/span><\/span><\/span><\/span><\/span><\/span>(<\/span>x<\/span>i<\/span><\/span><\/span>\u200b<\/span><\/span><\/span><\/span><\/span>)<\/span>\u2248<\/span><\/span>h<\/span>f<\/span>(<\/span>x<\/span>i<\/span><\/span>\u200b<\/span><\/span>)<\/span>\u2212<\/span>f<\/span>(<\/span>x<\/span>i<\/span>\u2212<\/span>1<\/span><\/span>\u200b<\/span><\/span>)<\/span><\/span>\u200b<\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/p>\n

For the second derivative:<\/p>\n

f\u2032\u2032(xi)\u2248f(xi)\u22122f(xi\u22121)+f(xi\u22122)h2f”(x_i) \\approx \\frac{f(x_i) – 2f(x_{i-1}) + f(x_{i-2})}{h^2}<\/span>f<\/span>\u2032\u2032<\/span><\/span><\/span><\/span><\/span><\/span><\/span>(<\/span>x<\/span>i<\/span><\/span><\/span>\u200b<\/span><\/span><\/span><\/span><\/span>)<\/span>\u2248<\/span><\/span>h<\/span>2<\/span><\/span><\/span><\/span>f<\/span>(<\/span>x<\/span>i<\/span><\/span>\u200b<\/span><\/span>)<\/span>\u2212<\/span>2f<\/span>(<\/span>x<\/span>i<\/span>\u2212<\/span>1<\/span><\/span>\u200b<\/span><\/span>)<\/span>+<\/span>f<\/span>(<\/span>x<\/span>i<\/span>\u2212<\/span>2<\/span><\/span>\u200b<\/span><\/span>)<\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/p>\n

\u00a0Example Calculation<\/strong><\/h3>\n

Given Data Points:<\/strong><\/h4>\n
\n\n\n\n\n\n
xx<\/span>x<\/span><\/span><\/span><\/span><\/th>\n1.0<\/th>\n1.2<\/th>\n1.4<\/th>\n<\/tr>\n<\/thead>\n
f(x)f(x)<\/span>f<\/span>(<\/span>x<\/span>)<\/span><\/span><\/span><\/span><\/td>\n2.718<\/td>\n3.320<\/td>\n4.055<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<\/div>\n

Step Size<\/strong>: h=0.2h = 0.2<\/span>h<\/span>=<\/span><\/span>0.2<\/span><\/span><\/span><\/span><\/h4>\n

Using the backward difference formula<\/strong>:<\/p>\n

f\u2032(1.4)\u2248f(1.4)\u2212f(1.2)hf'(1.4) \\approx \\frac{f(1.4) – f(1.2)}{h}<\/span>f<\/span>\u2032<\/span><\/span><\/span><\/span><\/span><\/span><\/span>(<\/span>1.4<\/span>)<\/span>\u2248<\/span><\/span>h<\/span>f<\/span>(<\/span>1.4)<\/span>\u2212<\/span>f<\/span>(<\/span>1.2)<\/span><\/span>\u200b<\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span> f\u2032(1.4)\u22484.055\u22123.3200.2=3.675f'(1.4) \\approx \\frac{4.055 – 3.320}{0.2} = 3.675<\/span>f<\/span>\u2032<\/span><\/span><\/span><\/span><\/span><\/span><\/span>(<\/span>1.4<\/span>)<\/span>\u2248<\/span><\/span>0.24.055\u2212<\/span>3.320<\/span>\u200b<\/span><\/span><\/span><\/span><\/span>=<\/span><\/span>3.675<\/span><\/span><\/span><\/span><\/span><\/p>\n

Final Answer:<\/strong> f\u2032(1.4)\u22483.675f'(1.4) \\approx 3.675<\/span>f<\/span>\u2032<\/span><\/span><\/span><\/span><\/span><\/span><\/span>(<\/span>1.4<\/span>)<\/span>\u2248<\/span><\/span>3.675<\/span><\/span><\/span><\/span><\/p>\n

\u00a0When to Use Backward Difference?<\/strong><\/h3>\n

\u00a0When you have data points and want to approximate derivatives using previous values.
\u00a0When solving numerical PDEs like heat equations<\/strong> and wave equations<\/strong>.
\u00a0In finite difference schemes<\/strong> for boundary conditions in numerical simulations.<\/p>\n

Would you like a Python code<\/strong> to implement this?<\/p>\n

Computer Science\/Numerical Methods\/ Finite difference method( backward difference)<\/a><\/h3>\n

18MAT21 Module 5 NUMERICAL METHOD CONTENTS: \u2022 …<\/a><\/h3>\n

B.Tech 4th Semester MATHEMATICS- …<\/a><\/h3>\n

The finite difference method<\/a><\/h3>\n

The Finite Difference Method (FDM)<\/strong> is a numerical technique used to approximate derivatives. One important type is the Backward Difference Method<\/strong>, which is especially useful for solving differential equations when data is given at the end of an interval<\/strong> or for time-stepping in numerical simulations.<\/p>\n

\n

\ud83d\udd39 Backward Difference Formula<\/strong><\/h2>\n

For a function f(x)f(x)<\/span>f<\/span>(<\/span>x<\/span>)<\/span><\/span><\/span><\/span>, the backward difference approximation<\/strong> of the first derivative is:<\/p>\n

f\u2032(xn)\u2248f(xn)\u2212f(xn\u22121)hf'(x_n) \\approx \\frac{f(x_n) – f(x_{n-1})}{h}<\/span>f<\/span>\u2032<\/span><\/span><\/span><\/span><\/span><\/span><\/span>(<\/span>x<\/span>n<\/span><\/span><\/span>\u200b<\/span><\/span><\/span><\/span><\/span>)<\/span>\u2248<\/span><\/span>h<\/span>f<\/span>(<\/span>x<\/span>n<\/span><\/span>\u200b<\/span><\/span>)<\/span>\u2212<\/span>f<\/span>(<\/span>x<\/span>n<\/span>\u2212<\/span>1<\/span><\/span>\u200b<\/span><\/span>)<\/span><\/span>\u200b<\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/p>\n

Where:<\/p>\n