<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=utf-8">
</head>
<body text="#000000" bgcolor="#FFFFFF">
That calculation I did quite some time ago. So, I cannot remember
my reasoning behind it. Since I didn't see an explanation for it in
Constraint_U.pdf and didn't know how Madsen and Pavel got the
weights, I likely just assumed the ε3d_dn would be the same as the
ε3d_up.<br>
<br>
I could be wrong, but I think the value for ":3dd001" and ":3d 001"
are from the "x lcore -up" [
<a class="moz-txt-link-freetext" href="https://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at/msg17160.html">https://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at/msg17160.html</a>
].<br>
<br>
So it might be that the weights (occupation numbers) come from
NiO.incup_+.50 and NiO.incup_+.5-1.<br>
<br>
<b>In Constraint_U.pdf</b><br>
<br>
NiO.incup_+.50 contains:<br>
<br>
N,KAPPA,OCCUP<br>
3, 2,<font color="#ff0000">4 <font color="#000000">3D*
or </font></font>:3dd001<br>
3,-3,<font color="#009900">5 </font><font
color="#ff0000"><font color="#000000">3D or </font></font>:3d 001<br>
<br>
NiO.incup_+.5-1 contains:<br>
<br>
3, 2,<font color="#ff0000"><font color="#000099">4</font> </font><br>
3,-3,<font color="#996633">5</font><br>
<br>
<blockquote type="cite"
cite="mid:CAHYowjGSs4gYJcc7kyhscTg8atd1CSkVY-UtNsYaA2Gfp+Kx1w@mail.gmail.com">
<div dir="ltr"><span id="gmail-result_box" class="gmail-"
lang="en"><span class="gmail-">I agree that the occupation for
spin-up (the addition of an electron) is 4 and 5</span></span><br>
ε3d_up = (<font color="#ff0000">4</font>*0.109629972+<font
color="#009900">5</font>*0.126221709)/<wbr>9 =
0.11884760366667<span id="gmail-result_box" class="gmail-"
lang="en"><span class="gmail-"><br>
<br>
but when we remove an electron for spin-Dn we have to put 4
and 3 or 4 and5?</span></span><br>
<br>
ε3d_dn = -(<font color="#000099">4</font>*-0.292400769+(<font
color="#996633">5</font>or3)*-0.<wbr>274004694)/ ...<br>
</div>
</blockquote>
<br>
</body>
</html>