ORCA做metaDynamics二维采样报错Error: More than 64 arguments (73)

Dempey · 发表于 Post on 2025-3-21 22:37:49

各位老师好，我最近想使用ORCA做AIMD的metaDynamics，现在用经典的SN2反应进行尝试，发现设置二维采样后就跑不通了，使用ORCA 6.0.1，输入文件和输出文件见附件，显示如下报错：

CSANParser::CreateImplicitCastSet(): Error: More than 64 arguments (73).
Abort during parsing at orca_md/san_parser_tools.cpp:3358 in void CSANParser::CreateImplicitCastSet(std::vector<CSANImplicitCastSet> &, const std::vector<CSANInputElement *> &, const std::vector<CSANType> &) const.

复制代码

metaD.inp (1.01 KB, 下载次数 Times of downloads: 6)

metaD.out (20.35 KB, 下载次数 Times of downloads: 2)

chands · 发表于 Post on 2025-4-13 22:40:00

我遇到同样的问题。算了好几个算例，如果算二维metadynamics就报错，但是删掉一个Colvar就没问题。不知道为什么。

我用了楼上的算例，除了没用并行之外其他没变，在超算上运行ORCA6.0.1

! wB97X-D3 TZVP RIJCOSX autoAux MD
%maxcore 2000
#%pal nprocs 10 end
%md
Timestep 0.5_fs
Thermostat CSVR 298.15_K TimeCon 100.0_fs
InitVel 298.15_K
Dump Position Stride 1 Format xyz Filename "trajectory.xyz"
Manage_Colvar Define 1 Distance Atom 0 Atom 4
Metadynamics Colvar 1 Wall Lower 1.0 50.0 Upper 30.0 50.0 Range 0.0 35.0 350
Manage_Colvar Define 2 Distance Atom 0 Atom 5
Metadynamics Colvar 2 Wall Lower 1.0 50.0 Upper 30.0 50.0 Range 0.0 35.0 350
Metadynamics HillSpawn 50 0.5 0.5 Store 500
Constraint Add Center 0
# MetaDynamics WellTempered 6000_K
# MetaDynamics Lagrange 100.0 10.0 200.0_K 10.0_fs
# Restart IfExists
Run 2000
end
* xyz -1 1
C -0.78767124 0.49657534 0.00000000
H -0.43099840 1.00097353 0.87365150
H -0.40717895 0.98370108 -0.87340376
H -1.85726636 0.52199613 -0.01483424
Cl -0.24087034 -1.17615690 0.02399276
Br -1.58973594 2.95019143 -0.03519333
*

复制代码

输出文件见附件，其中报错信息如下：

Handling functions, iteration 1...
ERROR : MD Program returns an error
cannot continue the md simulation
COMMAND : HOME/orca_6_0_1/orca_md -orca /HOME/MTD/metaD
RETURN CODE : 34304
[file orca_main/main_driver_md.cpp, line 190]:

复制代码

不知道RETURN CODE：30304，以及line 190是什么意思，请各位老师看看。

chands · 发表于 Post on 2025-4-14 22:09:10

本帖最后由 chands 于 2025-4-14 22:10 编辑

查了ORCA FORUM，上面有类似问题。
https://orcaforum.kofo.mpg.de/vi ... metadynamics#p44393
省流：ORCA出bug了，但是ORCA 6.0.1没有及时更新，可能在ORCA 6.1。当前解决办法，在%MD部分每一行结尾加分号（;）。亲测有用。

原文：
Re: Help for too many arguments in %md block
Unread post by Brehministrator » Sun Nov 03, 2024 2:53 am

That is an internal error of the input interpreter which was not expected to happen. There is no such thing as a limit on the argument count in the MD input. It is just a bug. I am sorry for the inconvenience.

The bug is already fixed since months in my local developer version, but as the changes were made together with the addition of a few new features, the bugfix could unfortunately not make it into ORCA 6.0.1, which is strictly a bugfix release without new features.

In the meantime, please just end each line with a semicolon (";"), it should help. The next feature release (maybe ORCA 6.1) will have this issue fixed.

Best regards,
Martin

P.S.: If you should be interested in the underlying cause: The MD input language is very permissive with respect to syntax. You can have multiple function calls in one line, and no semicolons or parentheses around the function arguments are required. Therefore, the parser figures out by itself which tokens are still arguments to the last function call, or which tokens are separate function calls on their own. During this check, there was a stupid bug with this arbitrary 64 argument limit. With the semicolons at the line endings, the parser can be sure that the function call is finished there and no more arguments will follow on the next line. Therefore this workaround works.

Dempey · 发表于 Post on 2025-4-15 15:48:31

chands 发表于 2025-4-14 22:09
查了ORCA FORUM，上面有类似问题。
https://orcaforum.kofo.mpg.de/vi ... metadynamics#p44393
省流：OR ...

谢谢咯，一直想去FORUM上找找，但还没行动

		自动登录 Automatic login	找回密码 Forget password
密码 Password			注册 Register

[ORCA] ORCA做metaDynamics二维采样报错Error: More than 64 arguments (73)

浏览过的版块