Tue, 14 Nov 2006

Ulrike Heiter [Subject: /Discussion/Documentation/Users] 

Full references in VALD extraction output:

At the Prague meeting (Aug. 2006), it was decided that Nik will
implement this into the extraction software, and Friedrich will provide
necessary input data (lists of references).

Tue, 18 Jul 2006

Christian Stütz [Subject: /Discussion/Documentation/Users] 


5) Update of the web-documentation and the ems text/latex
   documentation: I could do that
   Full reference strings in the new extraction software:
     Good idea. How about a reference list that can be used
     by the extraction software and display on the webpage.
     I would be interested to see something of the new 
     extraction software, too. How does it work, can we 
     have a peek, preview?

Fri, 30 Jun 2006

Tanya Ryabchikova [Subject: /Discussion/Documentation/Users] 

Dear frieds,
    Suppose, I need to extract 100 A for late A star, and depending on
spectral region I am getting more than 1000 lines. For each line we may 
have different sources for WL's, log(gf)'s, Lande (if available), damping 
constants. Do you estimate how many references user will get for such an 
extraction ? I am still not sure how to distribute reference field in the
case when we have more than one reference per line. Just couple of days 
ago I was creating a new list for Ca I with the experimental data only.
These lines have log(gf)'s from one paper and VdW constants from another
paper. What should I do ? Create separate file for VdW only as was done 
for Paul's calculations but for experimental data ?

                                        Regards, Tanya

Ulrike Heiter [Subject: /Discussion/Documentation/Users] 

Dear all,

Just a quick comment on the "References" topic:

The VALD-3 format includes a character field (length 7) which will
contain the abbreviation for the reference for each line (more details
later). We will try to recover these from the VALD-2 comment field
during the conversion - in part automatically and in part they will be
included in the lists which are being prepared by Tanya and Rimma.

Then the extraction software could be adapted to retrieve the full
reference for each line from a lookup table and include it in the
output. 

Best regards,
Ulrike

Friedrich Kupka [Subject: /Discussion/Documentation/Users] 

>>    - The lists of references for VALD data sets should be improved.
>>      a) Some references are missing in the lists on the website (e.g. the
>>         reference code "MFW").
>
>>      b) It should be made easier for users to see the original references
>>         directly after extraction from VALD.
>>         As it is now, the output of an extraction request contains the VALD
>>         dataset names, and in the case of "SHOWLINE" also the code for the
>>         reference in the comment field. So, after extraction, one has to look
>>         up the reference on two different pages on the website.
>>         The VALD dataset names could be replaced in the output by the full
>>         references.
>>
Yes. The first point is just an omission, the second needs some work. The 
problem is that so far there has not been any agreement about what should
be put into the comment fields. Moreover, as this information is usually
given in an irregular way, it requires lots of manual work. For sure, 
however, this will help to point back in VALD references in the most 
direct way to the original references (even those collected by Kurucz and
in the Bell lists).

Thu, 29 Jun 2006

Ulrike Heiter [Subject: /Discussion/Documentation/Users] 

5) Update VALD documentation for users
   
   - The documentation on the VALD website has to be updated to include
     information on VALD-3.

   - The lists of references for VALD data sets should be improved.
     a) Some references are missing in the lists on the website (e.g. the
        reference code "MFW").
     b) It should be made easier for users to see the original references
        directly after extraction from VALD.
        As it is now, the output of an extraction request contains the VALD
        dataset names, and in the case of "SHOWLINE" also the code for the
        reference in the comment field. So, after extraction, one has to look
        up the reference on two different pages on the website.
        The VALD dataset names could be replaced in the output by the full
        references.

Wed, 28 Jun 2006

Ulrike Heiter [Subject: /Discussion/Documentation/Users] 

To:      "Bengt Edvardsson"
	 "Anna Önehag"

Hi Bengt and Anna,

In the first VALD paper (Piskunov et al. 1995A&AS..112..525P), there
is a description of the VALD extraction software (Section 3),
including the control file and the ranking parameters (3.2 and 3.3).
There you find an explanation how data for one and the same line are
extracted from different files: within the wavelength window given in
the control file, lines must have the same ion, J-values and upper
energy level (to within 0.1%).

The second VALD paper is Kupka, et al. 1999A&AS..138..119K, with
further information on the database.

/Ulrike