https://bugs.documentfoundation.org/show_bug.cgi?id=165802
Bug ID: 165802
Summary: Linear Solver claims that "solving successfully
finished" for a non-linear problem but the 'solution'
violates a constraint
Product: LibreOffice
Version: 24.8.5.2 release
Hardware: All
OS: All
Status: UNCONFIRMED
Severity: normal
Priority: medium
Component: Calc
Assignee: [email protected]
Reporter: [email protected]
Description:
The attached workbook demonstrates an optimisation problem with a
non-linearity.
When using the "LibreOffice Linear Solver" on this problem, Calc reports that
"solving successfully finished". But the "solution" violates the constraint
that B14=0.
To be clear, the bug is not that a linear solver can't solve this problem. The
bug is that Calc reports success - what it should do is to detect that the
'solution' violates a constraint, and report that.
Steps to Reproduce:
Open the attached spreadsheet, and run the model with the given Solver settings
(i.e. maximise cell D12 by varying B10 and B11, subject to the constraint
B14=0, assuming the variables are non-negative, using “LibreOffice Linear
Solver”)
Actual Results:
Calc reports that “solving successfully finished”, but the “solution” violates
the constraint that B14=0.
Expected Results:
Calc should report that no acceptable solution can be found. (Due to the
non-linear dependence of B9 on B10, a linear algorithm is not expected to be
able to find a solution - that's not a problem. The problem is that the solver
reports that it has succeeded, when it hasn't.)
Reproducible: Always
User Profile Reset: No
Additional Info:
Version: 24.8.5.2 (X86_64)
Build ID: 480(Build:2)
CPU threads: 16; OS: Linux 6.12; UI render: default; VCL: kf6 (cairo+wayland)
Locale: en-CA (en_US.UTF-8); UI: en-US
Calc: threaded
--
You are receiving this mail because:
You are the assignee for the bug.
