This file implements exponential and logarithmic functions using Taylor series and ratio‑based approximations. The goal is to provide accurate and stable versions of exp(), natural logarithm (ln), base‑10 logarithm (log), base‑2 logarithm (log2) and arbitrary base logarithm (logB) without relying on Dart’s built‑in math functions.
double exp(num n)Computes e^n using a series expansion for small |n| and repeated squaring for larger values. Handles negative arguments by returning 1 / exp(-n). Throws an exception for NaN or infinite input.
double ln(num n)Computes the natural logarithm of n. Uses ratio‑based approximations for values near 1 and reduction by powers of 10 or 2 for large values. Returns negative infinity for zero, NaN for negative numbers and throws for NaN input.
double log(num n)Alias for log10(n). Uses ln(n) / ln(10) internally. Equivalent to ln but scaled by log base conversion. Returns negative infinity for zero, NaN for negative input.
double log2(num n)Computes the base‑2 logarithm using ln(n) / ln(2). Handles the same edge cases as ln().
double logB(num n, num base)Computes the logarithm of n with respect to an arbitrary base using ln(n) / ln(base). Throws if the base is not positive or equals 1, and handles the same edge cases as ln().
These functions are deterministic and avoid using dart:math’s log or exp, providing explicit control over algorithmic precision. They are used by other modules (e.g. curves and rounding) to build higher‑level behaviour.