ConvergenceOptions

ConvergenceOptions = object

Defined in: packages/math-analysis/src/convergence-options.ts:21

When an iterative routine should stop — close enough, too many tries, or cancelled.

Every iterative routine in this package takes the same shape, so a caller configures them one way rather than learning a new option object per function. Every field is optional; the defaults are chosen per routine and documented on it.

What tolerance is measured against is the one thing that necessarily differs — a bracket width, a relative update, a residual — so each function states it on its own parameter rather than here.

Example

  bisect(f, 0, 2, { tolerance: 1e-6, maxIterations: 50 });
  ```;

## See

[Convergence and iteration limits](https://anthochamp.pages.grenadine.io/node-essentials/topics/math/calculus/convergence/)

## Properties

### maxIterations?

> `optional` **maxIterations?**: `number`

Defined in: packages/math-analysis/src/convergence-options.ts:32

Upper bound on iterations, so an input that never converges still
terminates.

***

### signal?

> `optional` **signal?**: `AbortSignal` \| `null`

Defined in: packages/math-analysis/src/convergence-options.ts:38

Aborts the iteration. Checked once per iteration, so the cost is bounded by
`maxIterations` rather than by the work each iteration does.

***

### tolerance?

> `optional` **tolerance?**: `number`

Defined in: packages/math-analysis/src/convergence-options.ts:26

Threshold below which the iteration is considered converged. Each routine
documents the quantity it compares against this.