[Pw_forum] ev.x gives wrong results

[anyy zsj]

Thank you. But there is still a discrepancy for the following data set:
10.3      -54.420098694852939
10.4      -54.716135146323531
10.5      -54.894109787500000
10.7      -53.873787839705884
modified ev.x gives 10.46368, while the fitting is 10.52255.


On Wed, Jul 29, 2015 at 4:23 AM, Paolo Giannozzi <p.giannozzi at gmail.com>
wrote:

> Funny: one of the parameters of the fit is good for small unit cells, but
> too small for larger cells. In line 162 of PW/tools/ev.f90, replace
>       deltapar(1) = 0.1d0
> with something larger:
>       deltapar(1) = 1.0d0
> should be sufficient. Thank you for noticing this
>
> Paolo
>
> On Tue, Jul 28, 2015 at 6:06 PM, anyy zsj <zsjanyy at gmail.com> wrote:
>
>> This data set is actual calculated because I use a large supercell:
>> 10.1      -48.570860572794118
>> 10.3      -49.368013045588235
>> 10.5      -49.714550792647060
>> 10.7      -49.667482497058820
>> 10.9      -49.305953287500003
>> But the fitting results are not compatible.
>>
>> 2015-07-28 12:01 GMT-04:00 Paolo Giannozzi <p.giannozzi at gmail.com>:
>>
>>> Please show me a set of data *produced by actual calculations* that
>>> ev.x doesn't fit properly
>>>
>>> paolo
>>>
>>> On Tue, Jul 28, 2015 at 5:51 PM, anyy zsj <zsjanyy at gmail.com> wrote:
>>>
>>>> Thank you very much!
>>>> Sure it is angstrom. My concern is that with the same data, ev.x and
>>>> the manual fitting give different results as shown above. So I believe
>>>> something is there.
>>>> The discrepancy is gone if I use a small lattice parameter. For
>>>> example, if I scale the data by 1/3, both of them give the same results. So
>>>> what's going on?
>>>>
>>>> 2015-07-28 11:44 GMT-04:00 Paolo Giannozzi <p.giannozzi at gmail.com>:
>>>>
>>>>> On Tue, Jul 28, 2015 at 5:26 PM, anyy zsj <zsjanyy at gmail.com> wrote:
>>>>>
>>>>>
>>>>>>    1.4989396432   -6.0466776328
>>>>>>    1.5430233198   -6.0795705718
>>>>>>    1.5879629631   -6.0993455319
>>>>>>    1.6804430727   -5.9859764266
>>>>>> The first column is volume and the second energy with Ry. After ev.x
>>>>>> I input Ang, noncubic, 4, it gives me a results of 0.40 for minimal volume.
>>>>>> It is certainly wrong.
>>>>>>
>>>>>
>>>>> of course it is: ev.x assumes that the input data corresponds to a
>>>>> physical EOS. Your data isn't physical: the volume is way too small
>>>>>
>>>>>>
>>>>>> Another example for the below data:
>>>>>> 10.1      -48.570860572794118
>>>>>> 10.3      -49.368013045588235
>>>>>> 10.5      -49.714550792647060
>>>>>> 10.7      -49.667482497058820
>>>>>> 10.9      -49.305953287500003
>>>>>> The first column is lattice constant with angstrom and the second
>>>>>> energy with Ry.
>>>>>>
>>>>> After ev.x I input Ang, sc, 4, it gives me 10.50369, while I get a
>>>>>> different value 10.57517 by fitting to the formula in PRB 28,5480 (1983)
>>>>>> of Murnaghan's equation.:
>>>>>> http://gilgamesh.cheme.cmu.edu/doc/software/jacapo/appendices/appendix-eos.html
>>>>>>
>>>>>
>>>>>
>>>>> sure that the first column is in angstrom and not a.u.? This is what I
>>>>> get assuming it is in a.u.:
>>>>>
>>>>>  # equation of state: murnaghan.        chisq = 0.5402D-05
>>>>> # a0 = 10.5740 a.u., k0 =13776 kbar, dk0 =  4.17 d2k0 =  0.000 emin =
>>>>> -49.73649
>>>>> # a0 =  5.59554 Ang, k0 =1377.7 GPa,  V0 = ******* (a.u.)^3,  V0
>>>>> =175.197 A^3
>>>>>
>>>>> note the small value of the |chi|^2 (chisq) of the fit
>>>>>
>>>>> Paolo
>>>>>
>>>>> I have tested many groups of data but the discrepancy is there for
>>>>>> those large lattice parameters. Does ev.x use another equation or something
>>>>>> else I have missing? Thank you very much.
>>>>>>
>>>>>> _______________________________________________
>>>>>> Pw_forum mailing list
>>>>>> Pw_forum at pwscf.org
>>>>>> http://pwscf.org/mailman/listinfo/pw_forum
>>>>>>
>>>>>
>>>>>
>>>>> --
>>>>> Paolo Giannozzi, Dept. Chemistry&Physics&Environment,
>>>>> Univ. Udine, via delle Scienze 208, 33100 Udine, Italy
>>>>> Phone +39-0432-558216, fax +39-0432-558222
>>>>>
>>>>> _______________________________________________
>>>>> Pw_forum mailing list
>>>>> Pw_forum at pwscf.org
>>>>> http://pwscf.org/mailman/listinfo/pw_forum
>>>>>
>>>>
>>>>
>>>> _______________________________________________
>>>> Pw_forum mailing list
>>>> Pw_forum at pwscf.org
>>>> http://pwscf.org/mailman/listinfo/pw_forum
>>>>
>>>
>>>
>>>
>>> --
>>> Paolo Giannozzi, Dept. Chemistry&Physics&Environment,
>>> Univ. Udine, via delle Scienze 208, 33100 Udine, Italy
>>> Phone +39-0432-558216, fax +39-0432-558222
>>>
>>> _______________________________________________
>>> Pw_forum mailing list
>>> Pw_forum at pwscf.org
>>> http://pwscf.org/mailman/listinfo/pw_forum
>>>
>>
>>
>> _______________________________________________
>> Pw_forum mailing list
>> Pw_forum at pwscf.org
>> http://pwscf.org/mailman/listinfo/pw_forum
>>
>
>
>
> --
> Paolo Giannozzi, Dept. Chemistry&Physics&Environment,
> Univ. Udine, via delle Scienze 208, 33100 Udine, Italy
> Phone +39-0432-558216, fax +39-0432-558222
>
> _______________________________________________
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://pwscf.org/mailman/listinfo/pw_forum
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20150729/29beb585/attachment.html>