Multideterminant WF with an empty spin channel

[Matej Ditte]

Hello,

I am trying to run a multidet CSF WF for the H_2 triplet, where N_el_down=0

First, for the single-det case, I used the default output from the converter:

<particleset name="e" random="yes" randomsrc="ion0">
  <group name="u" size="2">
    <parameter name="charge">-1</parameter>
  </group>
  <group name="d" size="0">
    <parameter name="charge">-1</parameter>
  </group>
</particleset>

<slaterdeterminant>
  <determinant id="updet"     group="u" sposet="spo_ud" size="2"/>
  <determinant id="downdet"   group="d" sposet="spo_ud" size="0"/>
</slaterdeterminant>

This produced the following error:

Fatal Error. Aborting at Some 'group' XML element node doesn't contain a 'size' attribute! 'size = 0' is not allowed in the input. Make appropriate adjustments to the input or converter.

I was able to work around this by simply removing

<group name="d" size="0">
  <parameter name="charge">-1</parameter>
</group>

and

<determinant id="downdet"   group="d" sposet="spo_ud" size="0"/>
       
After that, the calculation runs correctly.

However, when I try the multidet version:

<multideterminant optimize="no" spo_up="spo_ud" spo_dn="spo_ud">
  <detlist size="6" type="CSF" nca="0" ncb="0" nea="2" neb="0" nstates="8" cutoff="1e-05">
    <csf id="CSFcoeff_0" exctLvl="2" coeff="0.999768" qchem_coeff="0.999768" occ="11000000">
      <det id="csf_0-0" coeff="1" alpha="11000000" beta="00000000"/>
    </csf>
    <csf id="CSFcoeff_1" exctLvl="2" coeff="0.015259" qchem_coeff="0.015259" occ="00010010">
      <det id="csf_1-0" coeff="-1" alpha="00010010" beta="00000000"/>
    </csf>
........................
  </detlist>
</multideterminant>
     
I get the following error:

Fatal Error. Aborting at In SlaterDetBuilder: SPOSet "" is not found. Expected for MultiSlaterDeterminant.

I tried:
-removing spo_dn="spo_ud"
-removing ncb="0" and neb="0"
-removing all beta="..."
-using the old WF format

Is there a way to run this, or is the multidet WF not implemented for an empty spin channel?

I attached a complete example reproducing the error. The h2.wfj.xml file also contains a commented-out single-det version.

Thank you for the help.

Best regards,
Matej

[Paul R. C. Kent]

Thanks Matej. I have made an issue on GitHub referencing your report at https://github.com/QMCPACK/qmcpack/issues/5953 . Hopefully there is a known workaround for the current code and someone will reply, but my guess is that when the input handling was written the code was simply too cautious.

[Paul R. C. Kent]

If the current code is too cautious or needs protections for the single spin channel case, this should be a relatively simple fix.

[Matej Ditte]

Hi Paul, 

Thanks for creating the GitHub issue. For the moment, I managed to make it work with:

      <multideterminant optimize="no" spo_0="spo_ud">
          <detlist size="6" type="DETS" nc0="0" ne0="2" nstates="8" cutoff="1e-20">
                  <ci coeff="0.999768"  occ0="11000000"/>
                  <ci coeff="-0.015259" occ0="00010010"/>
                  <ci coeff="-0.010487" occ0="00110000"/>
                  <ci coeff="-0.008384" occ0="00001001"/>
                  <ci coeff="-0.006391" occ0="10010000"/>
                  <ci coeff="0.003147"  occ0="00000101"/>
          </detlist>
      </multideterminant>

(thanks to  Ye Luo's comment on GitHub)