This is known as (specifically, the error reduces by a constant factor of ( 1/2 ) each iteration).
The maximum possible error after ( n ) iterations is half the interval length: numerical methods bicen maths
After 4 iterations: ( c = 1.5625 ), interval width = 0.0625, error ≤ 0.03125 < 0.05. Approximate root: (true root ≈ 1.52138 – error is within bound). 6. Advantages and Disadvantages | Advantages | Disadvantages | |------------|---------------| | Guaranteed convergence (if conditions met) | Slow convergence (linear only) | | Simple to understand and implement | Requires a sign change interval a priori | | Robust – works for non-differentiable functions | Cannot find roots of even multiplicity (e.g., ( f(x)=x^2 )) | | Error bound known at each step | Does not extend easily to systems of equations | 7. Comparison with Other Methods | Method | Convergence Rate | Requires Derivative? | Guaranteed? | |--------|----------------|----------------------|--------------| | Bisection | Linear (1/2 factor) | No | Yes | | Newton-Raphson | Quadratic | Yes | No (may diverge) | | Secant | Superlinear (~1.618) | No | No | This is known as (specifically, the error reduces
[ |r - c_n| \le \fracb_n - a_n2 = \fracb_0 - a_02^n+1 ] | Guaranteed
[ \fracb_0 - a_02^n+1 < \varepsilon \quad \Rightarrow \quad n > \frac\ln(b_0 - a_0) - \ln(\varepsilon)\ln(2) - 1 ]
To achieve a desired tolerance ( \varepsilon ), we need: