Repository navigation
Decline a substitution through a step its own node declines - #2356
Merged
Rafael-SOWNet merged 1 commit intoNov 3, 2026
Merged
Conversation
Since #2355 a floor or a ceil whose argument is real at the destination and complex just off it declines its own limit, and SolveBySubstitution, which reads an expression's value at the destination, answered the expression it is in anyway: x / floor(arcsin(x)) at 1 from the right was 1, the value at 1, where the floor all the way in is 1 - i and the quotient (1 + i)/2; at -1 from the left it was 1/2 for (2 - i)/5. 2.6.0 answers the same. The substitution declines such a step as it already declines one that sits on a jump. And #2355 asked whether an argument was real at the integer 10^6 on the way to an infinity, which for 1/x! built 10^6!: the limit of floor(1/x!) at +oo went from 93 ms to 3.9 s. Both checks now take the argument at the decimal 1e6 with the downcasting off, as limitcheck samples, and it is back to about 0.2 s. limitcheck at SEED=16 and at DEPTH=3 on SEED=18 and 19 changes only the two wrong limits, now unevaluated. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Codecov Report❌ Patch coverage is Additional details and impacted files@@ Coverage Diff @@
## master #2356 +/- ##
=======================================
Coverage 86.77% 86.77%
=======================================
Files 312 312
Lines 68777 68783 +6
Branches 20034 20037 +3
=======================================
+ Hits 59680 59686 +6
Misses 4910 4910
Partials 4187 4187
Flags with carried forward coverage won't be shown. Click here to find out more.
🚀 New features to boost your workflow:
|
Rafael-SOWNet
deleted the
a-substitution-declines-a-step-its-node-declines
branch
November 3, 2026 22:23
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
A follow-up to #2355. limitcheck found it at
DEPTH=3,SEED=19. No issue was filed.limitright(x / floor(arcsin(x)), x, 1)1, where the values tend to(1 + i)/2limitleft(x / floor(arcsin(x)), x, -1)1/2, where they tend to(2 - i)/5limit(floor(1/x!), x, +oo)SolveBySubstitutionreads an expression's value at the destination, and it answered the expression the step sits in anyway:x / floor(arcsin(1 + 1/t))att = +oois1 / floor(pi/2), which is 1. All the way in,floor(arcsin(x))is1 - iin the library's own evaluation. The substitution already declines a step that sits on a real jump, mirroring the node's own decline, and now it declines this one too.1/x!built 10^6!. Both checks now take the argument at the decimal 1e6 with the downcasting off, as limitcheck samples.limitcheck at
SEED=16, and atDEPTH=3onSEED=18and19, changes only the two wrong limits, now unevaluated. The other declined limits are master's exactly. 883 limit tests pass.Tests:
ASubstitutionDeclinesAStepItsNodeDeclinesTest, 7 cases. The two declines and the 2 s cap onfloor(1/x!)fail on master; the 4 real steps pass on both. The full suite, AOT and the nine harnesses pass. #2355's BREAKING-CHANGES entry gains the two rows, measured on 2.6.0.🤖 Generated with Claude Code