space groups and space group choice

77 views
Skip to first unread message

Victor Naden Robinson

unread,
May 27, 2025, 5:41:33 AMMay 27
to Calyp...@googlegroups.com
Dear calypso mailing,

I have two questions:

1) How is the initial space group for structure generation chosen (i.e. for generation 1, structure 1)? Is it completely random or based on the time_of_day? I notice that different independent searches starting at the same time sometimes give the same space group.

2) I am testing and searching to confirm an experimentally known phase, however the searches produce zero structures with the P3 (143) space group (after relaxation -), though does produce many results with space group P3_1 (144), but the lower energy P3 (144) structure is never found, not any with P3 symmetry at all. I thought this might be due to the volume or atom distances, but they seem ideal. 
Testing of a few thousand structures:
   1 (  19)      -6.58313      P3_1(144)
   1 (  30)      -6.58310      P3_1(144)
   1 (   5)      -6.58309      P3_2(145)
   1 (  37)      -6.58307    P3_221(154)
   1 (  19)      -6.58290    P3_221(154)
   1 (   1)      -6.58128      P3_1(144)
   1 (  30)      -6.57806      P3_1(144)
   1 (   5)      -6.57715      P3_1(144)
   1 (   5)      -6.57714      P3_1(144)
   1 (   5)      -6.57714      P3_1(144)
Yet none of these large number of structures are P3 (symmetry tolerance(s) m=0.01, m=0.1). Actually I should plot a histogram of the space group distribution, perhaps, before and after relaxation.
Is the P3 (144) space group, for example, only created in later calypso generations of searching and hence I am not making it?

I will experiment with searching only the 144 space group.

2*) Is there a way to view the space group chosen to the structure generation (i.e. before it is relaxed) in the calypso output?

Regards, Victor

wangzy.physics

unread,
May 27, 2025, 8:45:04 AMMay 27
to Calyp...@googlegroups.com, CALYPSODev
Dear Victor, 

Thank you for choosing CALYPSO for your research.

The initial space group is selected at random from the valid collection, so each independent search may yield different results. Here, “valid” means that the number of atoms must be compatible with the Wyckoff combination of target space group—in CALYPSO this is defined by the number_of_atom key, which specifies the primitive‐cell atom count. And you can find the space group information in results/struct.dat under the keyword SpaceGroup after you conduct one generation searching, or you can use spglib/phonopy to check the symmetry of initial structures.

Regarding your issue with generating P3 symmetry structures, I have tested it myself and successfully produced P3 structures. Please try the latest version of calypso.x, which is attached. If it still fails to generate the expected structures, please send me your working directory so that we can investigate the issue.

Best regards,

Zhenyu Wang
Key Laboratory of Material Simulation Methods and Software, Ministry of Education
College of Physics
Jilin University
People's Republic of China

---- Replied Message ----
From Victor Naden Robinson<victornad...@gmail.com>
Date 5/27/2025 17:41
To <Calyp...@googlegroups.com>
Subject [CALYPSO] space groups and space group choice
--
You received this message because you are subscribed to the Google Groups "CALYPSO" group.
To unsubscribe from this group and stop receiving emails from it, send an email to CalypsoCode...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/CalypsoCode/CAB10xRhxBr%3DycDfakqqUxL5GWPN14vtW%3DYULOUkiZ9CwuaveNw%40mail.gmail.com.
Victor.zip

296389682

unread,
May 27, 2025, 8:45:23 AMMay 27
to wangzy.physics

您的邮件已收到,我会尽快处理!

This is an automatic reply, confirming that your email was received. I will reply as soon as possible. Thank you!

Reply all
Reply to author
Forward
0 new messages