Skip to content

Reduce a power of a linear over a power of a quadratic a power at a time - #1892

Merged
Rafael-SOWNet merged 4 commits into
masterfrom
a-power-of-a-linear-over-a-power-of-a-quadratic-is-reduced
Oct 15, 2026
Merged

Rafael-SOWNet merged 4 commits into
masterfrom
a-power-of-a-linear-over-a-power-of-a-quadratic-is-reduced

Conversation

@Rafael-SOWNet

@Rafael-SOWNet Rafael-SOWNet commented Oct 15, 2026 •

Copy link
Copy Markdown
Member

Part of #718.

A power of a linear that is not whole, times a linear, over a whole power of a quadratic — L^m N / Q^n — is reduced one power of Q at a time (SolveAPowerOfALinearOverAPowerOfAQuadratic). Each step writes the numerator against Q' and Q by a 2×2 Cramer solve; what is left is the same shape with n one lower, and over Q^1 the integrator already answers it. The last integral is asked with its two coefficients named, so the solve that ends the chain sees atoms rather than compound coefficients.

Guards: the rule answers only the question asked; it declines a bare x to a positive power, imaginary coefficients, and a quadratic whose discriminant is a polynomial square (that one factors, and the factored route is the better answer).

Measured

Pocket of 7,020 Rubi rows, 5 s per row, with both arms on master 37b4798. The branch has merged #1891 since then, and the unit test and build pass on the merge:

master this branch
solved 6,939 6,948
wrong 0 0
timeout 30 20

All nine new answers held when each row was rerun alone. Two other rows read as moves in the pocket and were timeouts on both sides when rerun alone (1.2.1.4:1120, 1.3.1:568).

Row Integrand Now
1.2.1.2:2596 (d+e*x)^(7/2)/(a+b*x+c*x^2)^2 0.24 s
1.2.1.2:2601 1/((d+e*x)^(3/2)*(a+b*x+c*x^2)^2) 0.29 s
1.2.1.2:2603–2607 (d+e*x)^(k/2)/(a+b*x+c*x^2)^3, k = 7, 5, 3, 1, −1 0.3–0.6 s
1.2.1.3:1136 (A+B*x)/(x^(3/2)*(a+b*x+c*x^2)^2) 0.21 s
Welz 102 9.6 s (master: unsolved)

On the 6,939 rows that both solve, total time is 1,170 s against 1,530 s on master. Answers run from 1.2K to 7.1K characters. Exact checks at 50–2000 digits, with every symbol pinned, passed 168 of 168 over the targets, and 8 of 8 on Welz 102. The families 0–8 corpora have no losses. The gate passed, and the suite run in two halves passed 15,401 tests.

BREAKING-CHANGES.md records the new answers.

🤖 Generated with Claude Code

https://claude.ai/code/session_012sonx8iAspMiwRwokT1Ura

Rafael-SOWNet and others added 4 commits October 15, 2026 06:51
L^m N/Q^n, for m not whole and N linear, is (alpha + beta x) L^(m + 1) Q^(1 - n)
plus L^m W/Q^(n - 1) with W linear again, alpha and beta from two equations
that make N - (alpha + beta x)(1 - n) L Q' a multiple of Q; over the first
power the integrator is asked with the two coefficients named.

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_012sonx8iAspMiwRwokT1Ura
…top to the rules before

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_012sonx8iAspMiwRwokT1Ura
…duced gives

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_012sonx8iAspMiwRwokT1Ura
…-over-a-power-of-a-quadratic-is-reduced

# Conflicts:
#	Sources/AngouriMath/Functions/Continuous/Integration/IndefiniteIntegralSolver.cs
#	Sources/AngouriMath/Functions/Continuous/Integration/Integration.Definition.cs
@Rafael-SOWNet Rafael-SOWNet added this to the 2.6.0 milestone Oct 15, 2026
@Rafael-SOWNet
Rafael-SOWNet merged commit 34780ec into master Oct 15, 2026
34 checks passed
@github-project-automation github-project-automation Bot moved this from Todo to Done in AngouriMath Oct 15, 2026
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

Status: Done

Development

Successfully merging this pull request may close these issues.

1 participant