Thread: [cclib-devel] Is there something you need from us?

SourceForge Headquarters 225 Broadway Suite 1600 San Diego, CA 92101 +1 (858) 422-6466

Hello Chris,

Just wondering whether you're waiting for us to do something, or
whether you need any help to get started?

BTW, anyone going to the ACS in Boston? Me, at least.

Regards,
   Noel

Hi all,

Sorry for the delay, I was busy with other work and the problem was a
little more involved than I expected. I've committed a parser that reads
in coordinates and coefficients for closed shell species. It's not
working with CDA yet, so there are a few bugs to work out. In any case,
the job is part done. Open shell species will take a little more time.

I've added a new script called merge_turbo to combine the various
turbomole files into a single output file.

On an unrelated note, I've written a python script that reads in
Cartesian coordinates of a coordination complex and divides the atoms
into fragments automatically for use with CDA. It's not strictly a cclib
tool but I'd just as soon include it here as anywhere else.

Chris

-----Original Message-----
From: Noel O'Boyle [mailto:bao...@gm...] 
Sent: Tuesday, July 24, 2007 10:39 AM
To: Christopher Rowley
Cc: cclib-dev List; Adam Tenderholt
Subject: Is there something you need from us?

Hello Chris,

Just wondering whether you're waiting for us to do something, or
whether you need any help to get started?

BTW, anyone going to the ACS in Boston? Me, at least.

Regards,
   Noel

> Sorry for the delay, I was busy with other work and the problem was a
> little more involved than I expected. I've committed a parser that  
> reads
> in coordinates and coefficients for closed shell species. It's not
> working with CDA yet, so there are a few bugs to work out. In any  
> case,
> the job is part done. Open shell species will take a little more time.

Delays are quite normal, at least for me, as I usually only work on  
cclib/PyMOlyze when research is going slowly or I have a major itch  
to scratch.

Do you know why your parser doesn't work with CDA? Is it exiting with  
an error, or are the numbers wrong?

> I've added a new script called merge_turbo to combine the various
> turbomole files into a single output file.

Excellent. I'm glad you've decided on a way to handle turbomole  
calculations. At some point, you should run the standard dvb  
calculations using turbomol and add these to the turbomole svn branch.

> On an unrelated note, I've written a python script that reads in
> Cartesian coordinates of a coordination complex and divides the atoms
> into fragments automatically for use with CDA. It's not strictly a  
> cclib
> tool but I'd just as soon include it here as anywhere else.

I don't really understand what you mean. Shouldn't you have to run  
three calculations (ie. A-B, A, and B) for CDA? And if so, I don't  
see how a script could help setup these calculations.

Adam

Chris, can you upload some test files? I've created an example
directory structure that may be helpful with a directory per
calculation (unlike the other parsers, which have a file per
calculation). Until you upload a few test files, there's not a lot we
can do to help.

If you look at the test files available for the other parsers in the
"basicXXXXX" directory, you will see a number called dvb_something.
These are calculations on 1,4-divinylbenzene; a geometry optimization,
a single point calculation (with overlap matrix), an unrestricted
single point calculation (with overlap matrix), a frequency calc, a
TD-DFT, MPx calcs. Oh yeah, and it's useful to include the input file
too.

Once these are in place, we can add them into our test framework. From
then on, the aim is simply to make the parser pass all the tests (this
is a form of so-called test-driven development).

Noel

On 25/07/07, Christopher Rowley <cro...@uo...> wrote:
> Hi all,
>
> Sorry for the delay, I was busy with other work and the problem was a
> little more involved than I expected. I've committed a parser that reads
> in coordinates and coefficients for closed shell species. It's not
> working with CDA yet, so there are a few bugs to work out. In any case,
> the job is part done. Open shell species will take a little more time.
>
> I've added a new script called merge_turbo to combine the various
> turbomole files into a single output file.
>
> On an unrelated note, I've written a python script that reads in
> Cartesian coordinates of a coordination complex and divides the atoms
> into fragments automatically for use with CDA. It's not strictly a cclib
> tool but I'd just as soon include it here as anywhere else.
>
> Chris
>
> -----Original Message-----
> From: Noel O'Boyle [mailto:bao...@gm...]
> Sent: Tuesday, July 24, 2007 10:39 AM
> To: Christopher Rowley
> Cc: cclib-dev List; Adam Tenderholt
> Subject: Is there something you need from us?
>
> Hello Chris,
>
> Just wondering whether you're waiting for us to do something, or
> whether you need any help to get started?
>
> BTW, anyone going to the ACS in Boston? Me, at least.
>
> Regards,
>    Noel
>
>

Noel,

I've committed the optimization, single point, and vibrational frequency
calculations. One other thing I should mention is that turbomole doesn't
strictly have an input file. You run a specialized program (define) to
generate the control, coord, basis, auxbasis, and mos files. The
calculation is subsequently done by running one of the programs or
scripts that is part of turbomole. 

Also, I have not run the calculations with symmetry (although turbomole
can make use of symmetry if it is explicitly activated). My parser won't
work if the symmetry is anything other than c1 because I use the $closed
shells sections of the control file to calculate the number of homos.
The format of this section is different if symmetry is enabled, and this
will require a little thinking to fix.

Chris

-----Original Message-----
From: Noel O'Boyle [mailto:bao...@gm...] 
Sent: Monday, July 30, 2007 3:02 AM
To: Christopher Rowley
Cc: cclib-dev List; Adam Tenderholt
Subject: Re: Is there something you need from us?

Chris, can you upload some test files? I've created an example
directory structure that may be helpful with a directory per
calculation (unlike the other parsers, which have a file per
calculation). Until you upload a few test files, there's not a lot we
can do to help.

If you look at the test files available for the other parsers in the
"basicXXXXX" directory, you will see a number called dvb_something.
These are calculations on 1,4-divinylbenzene; a geometry optimization,
a single point calculation (with overlap matrix), an unrestricted
single point calculation (with overlap matrix), a frequency calc, a
TD-DFT, MPx calcs. Oh yeah, and it's useful to include the input file
too.

Once these are in place, we can add them into our test framework. From
then on, the aim is simply to make the parser pass all the tests (this
is a form of so-called test-driven development).

Noel

On 25/07/07, Christopher Rowley <cro...@uo...> wrote:
> Hi all,
>
> Sorry for the delay, I was busy with other work and the problem was a
> little more involved than I expected. I've committed a parser that
reads
> in coordinates and coefficients for closed shell species. It's not
> working with CDA yet, so there are a few bugs to work out. In any
case,
> the job is part done. Open shell species will take a little more time.
>
> I've added a new script called merge_turbo to combine the various
> turbomole files into a single output file.
>
> On an unrelated note, I've written a python script that reads in
> Cartesian coordinates of a coordination complex and divides the atoms
> into fragments automatically for use with CDA. It's not strictly a
cclib
> tool but I'd just as soon include it here as anywhere else.
>
> Chris
>
> -----Original Message-----
> From: Noel O'Boyle [mailto:bao...@gm...]
> Sent: Tuesday, July 24, 2007 10:39 AM
> To: Christopher Rowley
> Cc: cclib-dev List; Adam Tenderholt
> Subject: Is there something you need from us?
>
> Hello Chris,
>
> Just wondering whether you're waiting for us to do something, or
> whether you need any help to get started?
>
> BTW, anyone going to the ACS in Boston? Me, at least.
>
> Regards,
>    Noel
>
>

> Also, I have not run the calculations with symmetry (although  
> turbomole
> can make use of symmetry if it is explicitly activated). My parser  
> won't
> work if the symmetry is anything other than c1 because I use the  
> $closed
> shells sections of the control file to calculate the number of homos.
> The format of this section is different if symmetry is enabled, and  
> this
> will require a little thinking to fix.

This is actually something we've had to deal with for the ADF parser.  
The format where moenergies is parsed is different depending on  
whether there is symmetry, whether its restricted or unrestricted,  
and (I think), whether certain options are given.  Luckily, these  
sections are all given different headings, so it wasn't too hard to  
figure out what to do. Hopefully Turbomole does something similar.

Adam