headers
Thomas Hamelryck | 1 Jul 2002 09:31
Picon
Favicon

Re: Why don't we start the documentations as well?

[...]

>For example, Once somebody learned how to use
> BioPython and wants to use HMMs where does he go? 

The HMM module was not Happy Doc'ed for some reason. If it was, you would 
have a list of classes & methods etc. and a lot of documentation. I checked 
in the source code - it contains a lot of info. Simply applying Happy Doc 
would make that information readily available. I do not see any advantage in 
generating that info manually. Or am I missing something?

> Currently he has to go to the source code

No, you can check the API documentation generated with Happy Doc:

http://www.biopython.org/wiki/html/BioPython/BiopythonCode.html

> if we give him a reference he can simply look at the
> classes, the different methods and understand how it works.

Again, that list of classes and methods and a description of what they do and 
return can be generated automatically with Happy Doc. 

But maybe you are thinking of some kind of templated tutorial model? That 
would be useful IMO. In many cases a simple enumeration of classes & methods 
is not enough to use the package efficiently, so you need more.

Cheers,

-Thomas
(Continue reading)

Permalink | Reply | 3 comments |
headers
Thomas Hamelryck | 1 Jul 2002 10:18
Picon
Favicon

PDB again


I added the PDB module documentation to the tutorial.tex file under "Advanced"
- section 4.5.

---
Thomas Hamelryck      Vrije Universiteit Brussel (VUB)
Intitute for Molecular Biology            ULTR Department
Paardenstraat 65    1640 Sint-Gensius-Rode, Belgium
                http://ultr.vub.ac.be/~thomas
Permalink | Reply | 0 comments |
headers
Yair Benita | 1 Jul 2002 16:31
Picon

Re: Why don't we start the documentations as well?

I guess its also fine, I don't have a preference as long as we get started
on that. What do you think Jeff?
Yair

on 1/7/02 9:31, Thomas Hamelryck at thamelry <at> vub.ac.be wrote:

> [...]
> 
>> For example, Once somebody learned how to use
>> BioPython and wants to use HMMs where does he go?
> 
> The HMM module was not Happy Doc'ed for some reason. If it was, you would
> have a list of classes & methods etc. and a lot of documentation. I checked
> in the source code - it contains a lot of info. Simply applying Happy Doc
> would make that information readily available. I do not see any advantage in
> generating that info manually. Or am I missing something?
> 
>> Currently he has to go to the source code
> 
> No, you can check the API documentation generated with Happy Doc:
> 
> http://www.biopython.org/wiki/html/BioPython/BiopythonCode.html
> 
>> if we give him a reference he can simply look at the
>> classes, the different methods and understand how it works.
> 
> Again, that list of classes and methods and a description of what they do and
> return can be generated automatically with Happy Doc.
> 
> But maybe you are thinking of some kind of templated tutorial model? That
(Continue reading)

Permalink | Reply | 2 comments |
headers
Yair Benita | 1 Jul 2002 16:37
Picon

Re: PCR module

on 29/5/02 8:29, Thomas Sicheritz-Ponten at thomas <at> cbs.dtu.dk wrote:

> Ok, I am working on it!
> Bio/SeqUtils/__init__.py

So, how does it work? Should I send my scripts and you go through them and
sort them out or would you prefer to define a strategy and I can fit the
methods to that?

I also have many PCR utilities including the extraction of exons from a
given coding sequence (in case someone wants to use genomic DNA as a
template). Will this be included in the SeqUtils module?

Yair
--

-- 
Yair Benita
Pharmaceutical Proteomics
Utrecht University
Permalink | Reply | 0 comments |
headers
Brad Chapman | 1 Jul 2002 18:17
Picon

Re: Why don't we start the documentations as well?

Hey guys;

Thomas:
> >> For example, Once somebody learned how to use
> >> BioPython and wants to use HMMs where does he go?
> > 
> > The HMM module was not Happy Doc'ed for some reason. If it was, you would
> > have a list of classes & methods etc. and a lot of documentation. I checked
> > in the source code - it contains a lot of info. Simply applying Happy Doc
> > would make that information readily available. I do not see any advantage in
> > generating that info manually. Or am I missing something?

Ugh, yeah, the HappyDoc documentation was badly out of date. I just
updated everything so all the current classes should be included. If I
missed something, please let me know!

Yair:
> I guess its also fine, I don't have a preference as long as we get started
> on that. What do you think Jeff?

More documentation is always better so whatever you want to work on is
fine with me. The HappyDoc generated documentation is just a way to
extract the classes and methods, and to get the comments from the source
code; this way people can look at pretty HTML pages instead of the code
if they like to.

Personally, I wouldn't duplicate what HappyDoc gives, as that is a ton 
of pretty boring manual work and very tough to keep up to date by hand.
But, any other documentation you want to work on is great!
(Continue reading)

Permalink | Reply | 0 comments |
headers
Jeffrey Chang | 1 Jul 2002 20:00
Picon

Re: Why don't we start the documentations as well?

On Mon, Jul 01, 2002 at 04:31:27PM +0200, Yair Benita wrote:
> I guess its also fine, I don't have a preference as long as we get started
> on that. What do you think Jeff?
> Yair

I like it.  Either way is fine -- either beefing up the Happy doc
documentation or working on a separate reference manual.

Do you have a read/write CVS account, Yair?  If you would like one to
help with this, please send me you preferred account name and shell.

Jeff
Permalink | Reply | 0 comments |
headers
Thomas Hamelryck | 1 Jul 2002 21:28
Picon
Favicon

Re: Why don't we start the documentations as well?


[Brad, on documentation]

> Personally, I wouldn't duplicate what HappyDoc gives, as that is a ton
> of pretty boring manual work and very tough to keep up to date by hand.
> But, any other documentation you want to work on is great!

What about manpage-like documentation for every module? That would be useful.
These documents could have standard features like module name, author,
synopsis, general use, usage examples, known bugs & limitations,
implementation comments, related modules etc. That would complement the
HappyDoc docs in a very nice way.

Cheers,

---
Thomas Hamelryck      Vrije Universiteit Brussel (VUB)
Intitute for Molecular Biology           ULTR Department
Paardenstraat 65   1640 Sint-Gensius-Rode    Belgium
              http://ultr.vub.ac.be/~thomas
Permalink | Reply | 0 comments |
headers
Thomas Hamelryck | 1 Jul 2002 21:29
Picon
Favicon

Re: Why don't we start the documentations as well?

[Brad]

> Ugh, yeah, the HappyDoc documentation was badly out of date. I just
> updated everything so all the current classes should be included. If I
> missed something, please let me know!

Nice!

One question/remark though:

The structured text formatting rules at
http://www.python.org/sigs/doc-sig/stext.html say that:

<quote>
Special symbology is used to indicate special constructs:

* A paragraph that begins with a '-', '*', or 'o' is treated as an unordered
list (bullet) element. (Note: In list paragraphs it is not necessary to
separate each list item with a blank line.)
</quote>

This is supposed to apply to HappyDoc, but it doesn't happen, so you get
funny stuff like:

Arguments: o id - string, the id that will be used for the structure of o
filename - name of the PDB file

ie, there is no bulleted list.

Maybe there is some option missing when running HappyDoc so that the
(Continue reading)

Permalink | Reply | 501 comments |
headers
Thomas Hamelryck | 2 Jul 2002 14:39
Picon
Favicon

KD tree module


Added regression test. 

The KD tree module now does:

-Find all neighbors within radius r of given point p
-Find all point pairs within radius r of each other

It finds all bonds in the large ribosomal subunit (90000
atoms) in under 4 seconds. 

BTW, there a lot of Biopython tests that fail due to trivial reasons like:

test test_ParserSupport failed -- 
Writing: 'Ran 3 tests in 0.005s', expected: 'Ran 3 tests in 0.111s'

Cheers,

---
Thomas Hamelryck      Vrije Universiteit Brussel (VUB)
Intitute for Molecular Biology           ULTR Department
Paardenstraat 65   1640 Sint-Gensius-Rode    Belgium
              http://ultr.vub.ac.be/~thomas
Permalink | Reply | 6 comments |
headers
Jeffrey Chang | 3 Jul 2002 09:51
Picon

Re: KD tree module

On Tue, Jul 02, 2002 at 02:39:02PM +0200, Thomas Hamelryck wrote:
> BTW, there a lot of Biopython tests that fail due to trivial reasons like:
> 
> test test_ParserSupport failed -- 
> Writing: 'Ran 3 tests in 0.005s', expected: 'Ran 3 tests in 0.111s'

Hey, good catch.  It looks like there's some testing code here that
uses unittest.py, which writes out how long some code took.  This is
breaking from run to run.

Is someone responsible for unittest.py?  The easiest thing to do is to
remove the verbiage that describes how long the tests took.  The right
thing to do would be to fix the two regression testing systems, but I
suspect that would be too much work.  What do people think?

Jeff
Permalink | Reply | 5 comments |

Gmane