Version 3.1/News

In 2002 the first version of ESP (ELSA at the time) with some 11,000 electronic structures was created.
During 2002-2005 ESP/ELSA was expanded to ca. 20,000 materials. See publications for details.

26-12-12: It has been noted that in some rare cases the fermi level is not at zero in the plots (e.g. Mn sc and bcc). This will be fixed.
10-10-12: Major uppgrade. More details will follow.
11-05-12: About 1,000 high quality fermi surfaces added. See ESP-FS at

09-08-11: More than 10k electronic structure searches have been done from more than 1k unique sites!
26-04-11: Small gap materials back on-line.

27-08-09: Small fermi energy errors seen, in particular for small gap materials.
For these cases the top of VB is chosen as e_F.
24-08-09: Some LDA bandgaps fixed (figs. were still OK)
17-08-09: Hit max removed when no wild cards ("*") are used.
19-06-09: Midsummer Eve, Cheers!
18-04-09: Bash related error corrected (some .png's were not displayed).

12-12-08: Small php error corrected. The total- and Fermi energies are now printed...
01-11-08: Spin-polarized calculations converged (ca. 12,000 d-compounds).
23-10-08: Complete db update.
15-07-08: J. Comput. Matr. Sci. paper accepted.

13-11-07: Similar problems and fix for lanthanides and actinides.
12-11-07: Moving semi-core states to valence fix the problem. Several entries are recalculated.
02-10-07: We have seen large core leakage on some metal carbides.

2005: Om antalvstates>10 exco=1.
2004: Atomic_conf_mvstk for storre core...
2004: Skriver till elektronegativitet samt differenslista for NN av anion-cation elektronegativiteter.
2004: Fixade authnamnproblem (;) i *.inp.
Skriver ut atom_type_oxodation_number i *.inp
Borjar ater boka med fracocc. Introducerade fracocctab[nn]
fracocc 0 - ingen fracocc
fracocc 1 - fracocc
fracocc 2 - fracocc man kan losas exact utan supercell
fracocc 3 - fracocc som ej kan losas utan supercell eller approx
fracocc 10 - nagon occ>1.0...

2003/09/15: fixen under visade sig inte racka. i fix_coord_error2(x) skrivs nu istallet
for tal, braken ut. Tex 0.5->1/2 0.866->sqrt(3)/2, 0.577->1/sqrt(3), osv
Det samma galler for PLAT.
2003/09/04: Fixade koordinat avrundningsfel for kotanis CTRL. See fix_coord_error(x);
2003: Fixade lite smabuggar. Sok pa 031228.
2003: Found bug in spec.Si!!! Jag laste 2p istallet for 3p (10 vale- --> 4vale-)
2003: Fixed bug for As-Kr in periodiska2.txt (nothing crucial...)
2003: The slow behaviour was due to the usage of pointers for vorder...
2003: Found bug in the new code (re-used "n" in nested loops)... WORKS nicely!
2003: Solved the problem from yesterday by creating a file from compare_setup.inp.c which
contains the desired oreder for vname[]..
2003: When migrating to the new setup.inp I found disturbing problems which makes it impossible to
directly cp the new setup.inp, overwriting to old. The spec number seem to have changed in the
new code. I will put the new setup.inp as when the files differ in the spec numbering.
NOTE that this implies extra causion when using compare_setup.inp.
2003: Added uc info to setup.inp (red uc, NN, etc). Translated core pop from kappa not to nl notation.
2003: NN dist i setup.inp; om tva NN ar pa samma avstand skrivs den med lagst idnr till filen
(wills kod funkar sa).
2003: i nncellen maste man kora fran -nx..nx,, -lz..lz (n,m,l=1 alt 2,3...) annars blir rMT fel...(tex 0..2nx funkar ej!)

2002/11/30: Fixade bug med POTCAR (om atomerna ej kom i ordning blev det fel pots...)
2002/10/17: Inforde speciella KPOINT regler for vasp, se nedan
2002/10/13: KPOINTS for VASP grid.
2002/02/22: is created (rasmol_32BIT -xyz
2002/02/22: The VASP files INCAR, POSCAR, POTCAR and KPOINTS are created.
2002: 20->40 i joblist utskriften.
2002: adderade if-satser for att fa while loopen av .cif.icsd att ga snabbare.
2002: labbade med att ta bort tmp filerna och istllet jobba i minnet - sparade bara lite tid...
2002: #Position undetermined in .cif fixed (exco=10).
2002: vname[] problem fixade. Tog bort ".".
2002: 2002 versionen av ICSD ger en annorlunda cif-fil. Fixade skillnaderna... (sok 021031)
2002: Bug men frococc<1.0 for flera olika atomer over samma koordinater fixat (tex 43973_AsIn).
2002: Maximera avstanden vid frcocc. Fungerar och kollad!
NOTERA att minsta occ av en symmetri ar att. Dvs det som avrundas till 0 satts till 1 sa att
nagon ucred aldrig utelamnas.
2002: FIXADE sa att fracocc funkar! ucsitesymname[nn][1] - eller ar det redan gjort via sattet den
reducerade cellen viks upp? Ja, sa ar det. Transformationsloopen ligger innan nrreducloopen.
fracocc=10 tar bort matr.
Vilka fracocc som ska till Job_list kan styras direkt vid Job_list utskriften.
2002: Andrade tillbaks nft1,2,3 da benchmarks visade att ingen tidsvist fanns for ett 2^n grid.
2002: Definierar ucocc_facocc for styrning av vad som ska skrivas ut.
2002: Borjade med fracocc cellerna. Strategin ar att avrunda till narmaste
mojliga niva. Kan fta bli bra men ibland elt at skogen...
OBS fracocc>0 fixas vid utskriften till setup.inp. Orginal uc skrivs i Goodie stuff.
OBS maxinuc kollen gors innan fracocc reduceringen.
2002: Fixade nft1,2,3 sa att nuffran blir 2^n, n=1,2,3,... max=128 min=16.
2002: Rebecka 30!!! fixade in sitesym i setup.inp.

2001/04/25: fixed so that non spin pol. runs are memory redused (ncqpar and nsdpar).
2001: Worked on the different bravais lattices (testing) and on the frac. occ. problem.
2001: Journal info and goodie stuff to the end of .inp.
2001: Bug fix. When two different valences for the same atom, .inp was incorrect.
2001: Sort Job_list option added.
2001: "is" (include bravais lattice) added in Include-exclude.
2001: Bug fix. Correct NNname is printed in .inp.
2001: Reletivistic is now standard.
2001: The bravais lattices are now consistent with Bradley & Cracknell -
rotations of uc were needed...

2000/10/31: modified so that spin polarised runs can be performed
had to introduce spec.xx_SP, c.f. spec.xx
symt.inp is modified if SP (z-direction!, do not forget that in data files.)
data.base1 file added. Only important for SP runs.