Xpress fixes #3576
Xpress fixes #3576
Conversation
|
Thanks for your pull request! It looks like this may be your first contribution to a Google open source project. Before we can look at your pull request, you'll need to sign a Contributor License Agreement (CLA). View this failed invocation of the CLA check for more information. For the most up to date status, view the checks section at the bottom of the pull request. |
|
Thanks for the pr. You need to sign the CLA. |
|
Ping. I cannot accept the PR if you do not sign the CLA. Thanks |
Ack. I am working on this but it is not that easy. I first have to clarify some things with my employer. |
|
OK. Hope you can sort it out.
Laurent Perron | Operations Research | ***@***.*** | (33) 1 42 68 53
00
Le lun. 5 déc. 2022 à 10:03, Daniel Junglas ***@***.***> a
écrit :
… Ping.
I cannot accept the PR if you do not sign the CLA.
Thanks
Ack. I am working on this but it is not that easy. I first have to clarify
some things with my employer.
—
Reply to this email directly, view it on GitHub
<#3576 (comment)>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/ACUPL3OFFYFJBGD6MQBCMYLWLWVVTANCNFSM6AAAAAASPKNYCU>
.
You are receiving this because you commented.Message ID:
***@***.***>
|
We also support reading parameters from a file, even though Xpress does not support parameter files itself. The format for parameters is always NAME=VALUE. If set from a string, settings must be separated by ';'. If read from a file, there can be at most one setting per line.
At the moment all logging goes to the standard output. That should be enough for debugging.
We must always query the dimensions of the original model.
This can never be correct.
This is more future proof than removing all message callbacks.
That control on its own has no effect anyway. On the other hand, we now print through the message callback unless we are requested to be `quiet()`. `XPRS_OUTPUTLOG` could be set via `SetSolverSpecificParametersAsString()` if deemed necessary by the user.
This was requested by the reviewer. Note that we pass around the `XPRSprob` instance rather than passing the `XpressInterface` instance and then calling functions like `underlying_solver()` on it. That seemed more reasonable, especially in light of the existing `XPRSgetnumcols()` and friends.
This makes the `virtual` keyword obsolete and also the code more readable.
Instead of logging to `std::cout`, log to the appropriate channels.
|
Hello @lperron, |
|
Great news. This is merged in main. |
|
we use abseil log, but good idea.
Laurent Perron | Operations Research | ***@***.*** | (33) 1 42 68 53
00
Le mar. 25 avr. 2023 à 15:24, Florian Omnès ***@***.***> a
écrit :
… ***@***.**** commented on this pull request.
------------------------------
In ortools/linear_solver/xpress_interface.cc
<#3576 (comment)>:
> + if (!quiet())
+ XPRSaddcbmessage(mLp, cbmessage, &std::cout, 0);
Excellent ! GLog offer a LogSink virtual class
<https://github.com/google/glog/blob/674283420118bb919f83ceb3d9dee31ef43ff3aa/src/glog/logging.h.in#L1582>,
applications can override it and use AddLogSink to handle logs from GLog.
—
Reply to this email directly, view it on GitHub
<#3576 (comment)>, or
unsubscribe
<https://github.com/notifications/unsubscribe-auth/ACUPL3JT5GY4J2JQAP4AEBLXC7F7VANCNFSM6AAAAAASPKNYCU>
.
You are receiving this because you modified the open/close state.Message
ID: ***@***.***>
|
|
Hello @pet-mit, no, I did not see your work. Yes, I would like to discuss things. Let's get in contact. |
|
Hello @djunglas |
Mizux
added
Feature Request
This patches fixes a lot of different issues in the Xpress connector of Google OR tools.