…#185) (#806) Each of these asks for behaviour the library already has, and each is pinned by the example the proposal itself gives, so that closing it rests on a test rather than on a reading. #246 a logarithmic solver, converting logarithms to a polynomial where possible log(2, x) + log(2, x - 1) = 1 -> { 2 } ln(x)^2 - 3*ln(x) + 2 = 0 -> { e, e^2 } #214 a ^ f(x) + b ^ g(x) + ... = 0 2^x + 4^x = 6 -> { 1 } 3^(2x) - 4*3^x + 3 = 0 -> { 0, 1 } #185 collecting unlike terms by treating each as an unknown e^x + sin(x) + 2e^x + 2sin(x) -> 3*(sin(x) + e^x) Every root is substituted back into its equation. A solver that answers the right root alongside a wrong one is not solving the equation, and the printed set does not say which it has done -- which is the whole reason #115 and #272 were defects rather than gaps. Grouped into one commit rather than three, against the usual habit of a branch per issue, because there is no behaviour change to bisect: it is one test file recording a measurement about three issues. Test-only; 8 pass. Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com>

logarihtmic equation solver #246

Description

Activity

Metadata

Metadata

Assignees

Labels

Type

Projects

Milestone

Relationships

Development

Issue actions