Dear Paolo,

thank you very much for your fast reply. Your patch works fine!

Best regards,
Malte

Am 30.10.2017 um 17:38 schrieb Paolo Giannozzi:
On Mon, Oct 30, 2017 at 1:11 PM, Malte Sachs <[email protected] <mailto:[email protected]>> wrote:

    [...] in version 6.1/6.2 I get:

    / Starting magnetic structure//
    //     atomic species   magnetization//
    //        U              NaN/

Uranium is poisonous for people, Sodium Nitride for computer codes

    [...] the structure doesn't contain Uranium atoms


Maybe I know what is happening and why: try first to run the same input without U (that is: ntyp=1, no line with U PP). If it works, apply the following patch, try again with U. A change done just before v.6.1 wasn't checking for a case like yours (no atoms of a given type)

Index: PW/src/move_ions.f90
===================================================================
--- PW/src/move_ions.f90    (revision 13970)
+++ PW/src/move_ions.f90    (working copy)
@@ -474,9 +474,11 @@
            END IF
         END IF
      END DO
-     starting_magnetization(i) = starting_magnetization(i) / REAL(nt)
-     angle1(i) = angle1(i) / REAL(nt)
-     angle2(i) = angle2(i) / REAL(nt)
+     IF ( nt > 0 ) THEN
+        starting_magnetization(i) = starting_magnetization(i) / DBLE(nt)
+        angle1(i) = angle1(i) / DBLE(nt)
+        angle2(i) = angle2(i) / DBLE(nt)
+     END IF
   END DO
   DEALLOCATE ( r_loc, m_loc )

--
Paolo Giannozzi, Dip. Scienze Matematiche Informatiche e Fisiche,
Univ. Udine, via delle Scienze 208, 33100 Udine, Italy
Phone +39-0432-558216, fax +39-0432-558222



_______________________________________________
Pw_forum mailing list
[email protected]
http://pwscf.org/mailman/listinfo/pw_forum


--
Malte Sachs
Anorganische Chemie, Fluorchemie
Philipps-Universität Marburg
Hans-Meerwein-Straße 4
35032 Marburg (Paketpost: 35043 Marburg)
Tel.: +49 (0)6421 28 - 25 39 5
http://www.uni-marburg.de/fb15/ag-kraus/

Attachment: smime.p7s
Description: S/MIME Cryptographic Signature

_______________________________________________
Pw_forum mailing list
[email protected]
http://pwscf.org/mailman/listinfo/pw_forum

Reply via email to