Newton-Raphson Root Finder Calculator

Approximate real roots of a quadratic polynomial using Newton-Raphson iterations, tolerance, and initial guess controls.

Quick Facts

Core Formula
x_(n+1) = x_n - f(x_n) / f'(x_n)
Use this as a planning and validation aid, then confirm assumptions with your context.

Your Results

Calculated
Estimated Root
-
The root of ax² + bx + c closest to your guess
Iterations Used
-
Steps taken before the change fell below tolerance
Residual |f(x)|
-
How close f(x) is to zero at that root (smaller is better)
Convergence Status
-
Whether the method settled on a root

Ready

Press Calculate to see the root and how many steps it took.

Frequently Asked Questions

Why does Newton-Raphson converge so much faster than bisection?
Bisection only uses the sign of f(x) and halves the search interval each step (linear convergence). Newton-Raphson uses the slope (derivative) to jump much closer to the root each time, roughly doubling the number of correct digits per iteration when the guess is already close — this is called quadratic convergence.
What happens if I pick a starting guess exactly at the vertex of the parabola?
The derivative f′(x) = 2ax + b equals zero exactly at the vertex x = −b/2a, so the calculator's safeguard (it stops if |f′(x)| < 1e-12) will halt the iteration immediately rather than divide by zero, and the Convergence Status will read "Max iter / derivative stall."
Can this calculator find complex roots?
No — this version tracks a single real-valued guess and will stall or diverge if the quadratic has no real roots (when b² − 4ac < 0). For complex roots, use the discriminant to check first, then a complex-root solver.
How many iterations should I expect for a well-chosen starting guess?
For a quadratic with a reasonable starting guess, 4-8 iterations is typical to reach a tolerance of 1e-6, thanks to quadratic convergence. If Iterations Used is at or near your Max Iterations setting, the starting guess was likely too far from any root or too close to the vertex.

What it is and when to use it

The Newton-Raphson method is an iterative root-finding technique: starting from a guess x₀, it repeatedly follows the tangent line of the function down to where that tangent crosses zero, using that crossing point as the next, usually better, guess. For a differentiable function this converges very fast — quadratically, meaning the number of correct digits roughly doubles each iteration — provided the starting guess is reasonably close to a root and the derivative isn't near zero there.

Use this calculator whenever you need to solve a quadratic (or approximate root of any smooth function you can linearize this way) without factoring by hand, when you want to see how sensitive convergence is to the starting guess, or when teaching or reviewing numerical methods coursework. It's especially useful for equations that don't factor into nice integers, where the quadratic formula would otherwise be the only option.

Formula and method

For a function f(x) with derivative f′(x), each iteration computes:

xn+1 = xn − f(xn) / f′(xn)

Here f(x) = ax² + bx + c is the quadratic defined by your a, b, c inputs, and f′(x) = 2ax + b is its derivative. The calculator repeats this update until either the change between successive guesses drops below your tolerance (converged) or it reaches Max Iterations without converging, at which point it also checks whether the derivative got dangerously close to zero (a "derivative stall," which happens near a turning point of the parabola where the tangent line is nearly horizontal and the next guess can shoot far away).

Worked example

With the calculator's own defaults — a=1, b=−5, c=6, initial guess x₀=3.5, tolerance 0.000001, max 20 iterations — the target function is f(x) = x² − 5x + 6 = (x−2)(x−3), which has two real roots at x=2 and x=3.

  • Iteration 1: x = 3.5 − 0.75/2 = 3.125
  • Iteration 2: x = 3.125 − 0.140625/1.25 = 3.0125
  • Iteration 3: x = 3.0125 − 0.012656/1.025 ≈ 3.000152
  • Iteration 4: x ≈ 3.0000000488
  • Iteration 5: the change from the previous guess drops below the 0.000001 tolerance, so it stops

Final result: Estimated Root = 3.00000000, Iterations Used = 5, Residual |f(x)| ≈ 0.000e+0, Convergence Status = Converged — matching what the calculator returns for those exact inputs. Note it converges to x=3, not x=2, because the starting guess 3.5 sits closer to (and on the same side of the vertex as) the root at 3.

Common mistakes and how to interpret results

  • Starting too close to a local extremum. If x₀ lands where f′(x) ≈ 0 (the parabola's vertex, at x = −b/2a), the tangent line is nearly flat and the next guess can jump far from any root or trigger the "derivative stall" status shown as Convergence Status.
  • Assuming it always finds every root. Newton-Raphson finds one root per run, whichever one the starting guess happens to converge toward. A quadratic has up to two real roots — try a second starting guess on the opposite side of the vertex to find the other one.
  • Setting tolerance too loose or too tight. A tolerance like 0.1 can report "converged" many iterations before the answer is actually precise; an extremely tight tolerance (below floating-point precision, roughly 1e-15) may never be satisfied and will run to Max Iterations instead.
  • Ignoring a non-zero residual. Iterations Used stopping doesn't guarantee a near-zero Residual |f(x)| — if the method hit "Max iter / derivative stall" rather than "Converged," check the residual value directly before trusting the root.

Related calculators

Related calculators