Apologies for the confusion - the "?" indicated I didn't know the
answer, rather than didn't understand the question. I'm hoping the
others can weigh in on them.
On 12 November 2010 00:21, Andrew Dalke <da...@dalkescientific.com> wrote:
> On Nov 11, 2010, at 9:58 PM, Noel O'Boyle replied to Andrew Dalke:
>>> Since I'm not (directly) calling WriteMolecule and since my
>>> test cases don't support the --gen2D or --gen3d flags, this
>>> warning isn't that helpful in a library function.
>>
>> This is by design (added recently). Up until recently, new users
>> thought that Open Babel didn't work, and complained to the mailing
>> list. I remember it confused me too when I first did a SMILES to SDF
>> conversion and it came up all zeros! This way we annoy you and me but
>> everyone else is happy.
>
> I understand the need for the warning. It's useful and appropriate.
>
> But in a general-purpose library, this warning is less useful
> and rarely appropriate. It indicates that the library should have
> some way to flag to the caller that some action should be done,
> and let the caller figure out the right way to handle that.
>
> For example, in a GUI without a console, that message isn't
> helpful.
>
> However, as the number of people who have a problem with this
> are very small in number, I am not suggesting a change to the
> API. I wanted to point this out only if there was a simple
> solution which addressed everyone's problems.
>
>>> 3) Why does the fingerprint GetBit require a fingerprinter instance?
>>> That is, I have to go through a bound method
>>>
>>> x = ob.OBFingerprint.FindFingerprint("FP2")
>>> self.assertEquals(x.GetBit(v, 0), True)
>>>
>>> instead of using a static/global function
>>>
>>> self.assertEquals(ob.OBFingerprint.GetBit(v, 0), True)
>>
>> ?
>
> That is, why do I need the "FP2" or to load any fingerprinter.
> Is there any way to call GetBit otherwise?
>
>> All Begin/End methods do not work from Python. Many of them are
>> disabled from Python.
>
> Okay. Didn't know that.
>
>>> 6) I included tests for iterating over atoms and bonds in
>>> a molecule, and for atoms and bonds connected to an atom. I
>>> expected I could also iterate over the atoms in a bond, but
>>> there is no OBBondAtomIter . While not important, it appeals
>>> to my sense of symmetry and it's very simple to write.
>>
>> True, but will have to wait until next API change.
>
> What does "API" change mean here? This is part of the adapter
> layer to make Python work with OpenBabel. ... Ahh, you mean
> that the C++ layer doesn't have an iterator interface for
> this case, so the Python one won't have something like
>
> def OBBondAtomIter(bond):
> yield bond.GetBeginAtom()
> yield bond.GetEndAtom()
>
> until the C++ one does. Well, I can understand the reasoning
> behind keeping things in synch.
>
>
>>
>>> 7) Why can't I do mol.AddBond(C, N, 3) where C and N are atoms?
>>> Instead, I need to do mol.AddBond(C.GetIdx(), N.GetIdx(), 3)
>>
>> No reason I guess. Could cause problems where C and N are not members
>> of the molecule though.
>
> Which lead me to wonder why
>
>>>> import openbabel as ob
>>>> mol = ob.OBMol()
>>>> mol.AddBond(123,435, 3)
> True
>
> returned True. Then digging through the code I discovered
> "OBGenericDataType::VirtualBondData".
>
> which lead me to try
>
>>>> mol.AddBond(-123, -532, 3)
> True
>
> I don't intend for the code to change, only pointing out
> something I thought was interesting.
>
>>> 9) How does OpenBabel define "implicit hydrogen"?
>
>> ?
>
> Take away the other things I said. It was my attempt at
> figuring out what OB does. I think I know the answer but
> it's different from RDKit, OEChem, and Daylight, and
> not what I'm used to, so I'm looking for enlightenment.
>
>> I think no C arrays are accessible (I think most instances can be
>> worked around). You need to use atom.GetVector().
>
> Through the magic of ctypes. I use it in my fingerprint
> reader, for example, to get faster access to the actual
> fingerprint data.
>
>>> 15) In an OBRing, if there are multiple non-C atoms, which one
>>> is the root? And since ring.GetRootAtom() returns an integer,
>>> shouldn't it be named GetRootAtomIdx()?
>>
>> ?
>
> What does the "?" refer to? There are two parts here.
>
> - "the root"
>
> mol = parse_smiles("C12CNCC3C1.C2CCC3")
> sssr = mol.GetSSSR()
> ring = sssr[0]
> idx = ring.GetRootAtom()
>
> The documentation for GetRootAtom:
>
> the index for the root atom. O for furan, S for thiazole,
> N for pyrrole. For 6 membered aromatic rings, the first
> non carbon atom is used for root. For 5 members rings the
> O, S or N (BOSum=3, valence=3) will be used for root
>
> That isn't enough information to figure out what it does.
> What is "first"? What if there are multiple non-carbons, as in
> "C1CCOCOCCNC1" -- which is first?
>
> - "GetRootAtomIdx()"
>
> The other API calls which end "*Atom()" return an OBAtom,
> while *Idx() returns an integer. This is an API call which
> returns an integer index instead of an actual atom.
>
>>> 19) I see that MMFF94's ValidateGradients() dumps output
>>> to stdout. The other ValidateGradients I tested (gaff and
>>> ghemical) do not do that, and the output probably isn't
>>> appropriate for a library.
>>>
>>> Then again, the API docs say that it's meant for debugging.
>>> Should the debugging API be published for general use?
>>
>> ?
>
> Who is the target audience for ValidateGradients()? Is
> it used for anything other than debugging?
>
>
> Andrew
> da...@dalkescientific.com
>
>
>
> ------------------------------------------------------------------------------
> Centralized Desktop Delivery: Dell and VMware Reference Architecture
> Simplifying enterprise desktop deployment and management using
> Dell EqualLogic storage and VMware View: A highly scalable, end-to-end
> client virtualization framework. Read more!
> http://p.sf.net/sfu/dell-eql-dev2dev
> _______________________________________________
> OpenBabel-Devel mailing list
> OpenBabel-Devel@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/openbabel-devel
>
------------------------------------------------------------------------------
Centralized Desktop Delivery: Dell and VMware Reference Architecture
Simplifying enterprise desktop deployment and management using
Dell EqualLogic storage and VMware View: A highly scalable, end-to-end
client virtualization framework. Read more!
http://p.sf.net/sfu/dell-eql-dev2dev
_______________________________________________
OpenBabel-Devel mailing list
OpenBabel-Devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/openbabel-devel
