The CASINO forum

Hi,
I need help figuring out a problem with jastrow optimization I am facing with positronic molecules.
My test system is Mg:e+. (no pseudopotential).
I realise I have to use the H term of the jastrow factor, but there is no documentation for it. The H term also requires the custom_striplet_dep block. I can’t figure how how to specify the spin-triplet groupings. Some help figuring this out would be greatly appreciated.
Abhijeet

Dear Abhijeet,

Here is a sample three-body H term (to go in correlation.data) using custom spin-dependence "-1":

START H TERM
Number of sets
1
START SET 1
Spherical harmonic l,m
0 0
Expansion order N_h
2
Spin dep (0->uu=dd=ud; 1->uu=dd/=ud; 2->uu/=dd/=ud)
-1
Cutoff (a.u.) ; Optimizable (0=NO; 1=YES)
7.0000000000000000 0
Parameter values ; Optimizable (0=NO; 1=YES)
-6.7299194156029003E-005 1 ! h_1,1,1,1
-7.8947577396053096E-004 1 ! h_1,2,1,1
-3.8568919150530520E-004 1 ! h_2,2,1,1
8.6020805550366698E-004 1 ! h_0,0,2,1
2.5189828771633071E-004 1 ! h_0,1,2,1
7.9415629418061844E-006 1 ! h_1,1,2,1
-5.6058048660126592E-005 1 ! h_1,2,2,1
-4.2736878055053135E-004 1 ! h_2,2,2,1
-6.9782720732719707E-003 1 ! h_1,1,1,2
3.5564660827365840E-003 1 ! h_1,2,1,2
-3.6816514441010638E-003 1 ! h_2,2,1,2
-1.3050240431361793E-002 1 ! h_0,0,2,2
-1.8583984431299379E-003 1 ! h_0,1,2,2
-1.1514227089071317E-004 1 ! h_1,1,2,2
-1.4524968182283329E-003 1 ! h_1,2,2,2
3.4177452972667870E-003 1 ! h_2,2,2,2
0.0000000000000000 0 ! h_1,1,1,3
0.0000000000000000 0 ! h_0,0,2,3
0.0000000000000000 0 ! h_1,1,2,3
0.0000000000000000 0 ! h_1,2,2,3
0.0000000000000000 0 ! h_2,2,2,3
0.0000000000000000 0 ! h_1,1,1,4
0.0000000000000000 0 ! h_1,2,1,4
0.0000000000000000 0 ! h_2,2,1,4
0.0000000000000000 0 ! h_0,0,2,4
0.0000000000000000 0 ! h_0,1,2,4
0.0000000000000000 0 ! h_1,1,2,4
0.0000000000000000 0 ! h_1,2,2,4
0.0000000000000000 0 ! h_2,2,2,4
END SET 1
END H TERM

H is a polynomial three-body electron-electron-electron term, a bit like the electron-electron-nucleus term f.

Here's the input block that defines spin-dependence -1:

%block custom_striplet_dep
no_striplet_deps 1
striplet_dep -1 4
1=2-3,1=2-4
1-3=4,2-3=4
1=1=1,2=2=2,3=3=3,4=4=4
1=1-2,1=1-3,1=1-4,1-2=2,1-3=3,1-4=4,2=2-3,2=2-4,2-3=3,2-4=4,3=3-4,3-4=4
%endblock custom_striplet_dep

This means that we are defining one custom triplet dependence, "-1". This splits the triplets of particles into 4 groups. The first has triplets involving particle types 1, 2 and 3 and triplets involving particle types 1, 2 and 4. (E.g., opposite-spin electrons (particle types 1 and 2) and a spin-up hole (particle type 3), and opposite-spin electrons with a spin-down hole (particle type 4). The "=" sign shows that the paring between an up-spin electron and the hole is the same as the pairing between the down-spin electron and the hole.

Type custom_striplet_dep for more information. Note that every triplet has to be listed, even if it is not used. Hence, in your case, you will have positron=positron=positron terms, whose coefficients can just be fixed.

Hope that helps. Best wishes,

Neil.

Sorry, that should of course be: type "casinohelp custom_striplet_dep".

Best wishes,

Neil.

abd43 wrote: I realise I have to use the H term of the jastrow factor, but there is no documentation for it. The H term also requires the custom_striplet_dep block. I can’t figure how how to specify the spin-triplet groupings. Some help figuring this out would be appreciated.

The H term is defined on p. 66 of the current manual.

The custom_striplet_dep block is defined both in casinohelp (as Neil says) and in the manual.

If this isn't the case in your version then you must be working with an out-of-date distribution..

Cheers
Mike