summary |
shortlog | log |
commit |
commitdiff |
tree
first ⋅ prev ⋅ next
Diane Trout [Sat, 25 Feb 2006 01:37:09 +0000 (01:37 +0000)]
tell doxygen to create docs for the qt classes too
Diane Trout [Sat, 25 Feb 2006 00:41:18 +0000 (00:41 +0000)]
skeleten of a qt version of mussa
One annoying thing about qmake is that each .pro file can only make one
binary. So the old make file which build the fltk gui and more importantly
the boost based unittests is now available as Makefile.noqt.
qmake-qt4 mussagl.pro will generate the qt4 makefile for building the qt4
mussa.
(And I may switch to boost build or scons so I can have one build file
for all the versions).
Diane Trout [Fri, 24 Feb 2006 07:17:16 +0000 (07:17 +0000)]
refactor nway_other.cxx
In order to understand how the transitive and radial nway comparisons
worked I wanted to make the algorithms fit in a single window. I did
this by taking various types of initialization loops that were duplicated
between both algortihms and refactored them into seperate functions.
(mostly these common loops handled building paths that were between
our various iterators on each species lists of FLPs)
I thought it might be nice to turn the various functions I ripped out
into a class that manages the march of the iterators across the lists
of FLPs.
I also added some fairly basic unit tests for the nway (transitive) comparison.
Diane Trout [Fri, 24 Feb 2006 01:08:54 +0000 (01:08 +0000)]
use exceptions to indicate Mussa::analyze failed
also modified the fltk gui code to report the exception text.
Diane Trout [Fri, 24 Feb 2006 00:47:32 +0000 (00:47 +0000)]
subanalysis menu option wont crash mussa
Diane Trout [Fri, 24 Feb 2006 00:36:01 +0000 (00:36 +0000)]
only include headers by names relative to project base path
Diane Trout [Thu, 23 Feb 2006 22:34:34 +0000 (22:34 +0000)]
build mussa with CXXFLAGS, so I can run gdb.
Diane Trout [Thu, 23 Feb 2006 06:58:03 +0000 (06:58 +0000)]
fixing the number of sequences at setup is silly
there's little reason to store seq_num as it's just as easily computed
from the size of our sequence vector.
Diane Trout [Thu, 23 Feb 2006 06:50:04 +0000 (06:50 +0000)]
fix inability to do reverse compliments
When I added the dynamic allocation of the all_matches vector, I assumed
that seqcomp would only be called once. It turns out that in actuality,
its called twice. Once for the forward compliment and once for the reverse
compliment. This includes a unittest that detects the problem and a fix
for the bug.
Diane Trout [Thu, 23 Feb 2006 02:23:23 +0000 (02:23 +0000)]
more thoroughly document FLPs
Diane Trout [Thu, 23 Feb 2006 02:15:10 +0000 (02:15 +0000)]
use boost to delete my temp file
the unittest now depends on boost/filesystem
Diane Trout [Thu, 23 Feb 2006 02:06:27 +0000 (02:06 +0000)]
refactor and test FLPs
FLPs now dynamically allocate the list of all_matches, which
has the bonus of not needing to pass in the size of the two
sequences that will be seqcomped.
I renamed win_num to size to make it more constent with other container
classes.
Also file_load and file_save were renamed to load and save, again
to be more consistent.
Lastly, I do make sure that I can construct FLPs and save and load them.
Diane Trout [Wed, 22 Feb 2006 20:16:21 +0000 (20:16 +0000)]
rename mussa_nway to nway_paths
Diane Trout [Wed, 22 Feb 2006 20:14:58 +0000 (20:14 +0000)]
i think a binary called unittests is more obvious.
Diane Trout [Wed, 22 Feb 2006 20:06:45 +0000 (20:06 +0000)]
rename the_Seqs to the_seqs
Diane Trout [Wed, 22 Feb 2006 08:27:24 +0000 (08:27 +0000)]
load a mupa file and test it
Start trying to isolate ConnWindow from the mussa_class object by using
accessors.
Try to get rid of seq_num and make mussa_class a bit more dynamic,
unfortunately when I killed a bit more in mussa_class than what's gone in
this patch the test cases noticed trying to free an invalid pointer.
Diane Trout [Wed, 22 Feb 2006 07:41:44 +0000 (07:41 +0000)]
add get_analysis_mode_name
get_analysis_mode_name returns a string representation of our analysis
name.
also added some consts to a couple of member functions
Diane Trout [Wed, 22 Feb 2006 07:04:44 +0000 (07:04 +0000)]
refactor and test sequence class
Diane Trout [Wed, 22 Feb 2006 06:42:18 +0000 (06:42 +0000)]
put exception definitions in one place and use them
use exceptions that are derived from runtime_error, and
use them for the cmdline argument parsing and sequence loading
Diane Trout [Wed, 22 Feb 2006 04:05:39 +0000 (04:05 +0000)]
always use all the prequisites when building
$? is the prereqs "newer" than the target
$^ is all the prerequisites
Needless to say when building binaries you need all the libraries,
even when they haven't changed.
Diane Trout [Wed, 22 Feb 2006 02:20:05 +0000 (02:20 +0000)]
force doxygen to look at all of our directories
Diane Trout [Wed, 22 Feb 2006 02:16:42 +0000 (02:16 +0000)]
rename some mussa_class members (ana_name to analysis_name)
also add some getters, reformat the parse_args if/else tree.
Change using character flags for analysis mode to using an enumeration.
Diane Trout [Wed, 22 Feb 2006 02:06:03 +0000 (02:06 +0000)]
test mussa class
Also the test case extension should match that of the other source
files.
Diane Trout [Wed, 22 Feb 2006 00:12:57 +0000 (00:12 +0000)]
unit test skeleton
Now we build a unit test binary, though currently there aren't any tests.
Diane Trout [Wed, 22 Feb 2006 00:07:33 +0000 (00:07 +0000)]
remove library blessings
Diane Trout [Tue, 21 Feb 2006 23:51:14 +0000 (23:51 +0000)]
break algorithm out into a subdirectory.
Diane Trout [Tue, 21 Feb 2006 23:01:44 +0000 (23:01 +0000)]
build fltk gui stuff as a library
Diane Trout [Tue, 21 Feb 2006 08:45:32 +0000 (08:45 +0000)]
we also want docs about the fltk gui code
Diane Trout [Tue, 21 Feb 2006 08:42:03 +0000 (08:42 +0000)]
add notes of what i learned about seqcomp and nway
stored notes as doxygen comments
Diane Trout [Sat, 4 Feb 2006 18:08:07 +0000 (18:08 +0000)]
add parse error exceptions for the command line args.
Diane Trout [Sat, 4 Feb 2006 18:07:25 +0000 (18:07 +0000)]
number of seqs is more informative than seq num
Diane Trout [Sat, 4 Feb 2006 00:38:24 +0000 (00:38 +0000)]
massive code move
mussa_gui_ was redundant so i moved all of those modules into
gui/
Also I used the non-recursive gnu make style build system.
And lastly there are some Doxygen compatible bits of doucmentation thrown in
here.
Diane Trout [Fri, 3 Feb 2006 02:30:04 +0000 (02:30 +0000)]
make headers less interdependent
The mussa headers all depended on including each other to get
all of the right include files. This has been broken and files are more
explicit at including what they depend on.
Also based on Accellerated C++, I followed the convention that header files
should never do a using namespace std, on the off chance that someone
wants to redefine something in the std namespace.
Diane Trout [Thu, 2 Feb 2006 23:16:38 +0000 (23:16 +0000)]
remove unnecessary member variable length.
One should trust the string.length() instead
Diane Trout [Tue, 27 Sep 2005 19:32:50 +0000 (19:32 +0000)]
[project @ 16]
Build settings for mingw32 (under cygwin)
Diane Trout [Tue, 27 Sep 2005 00:59:56 +0000 (00:59 +0000)]
[project @ 15]
As mussa is a c++ program switch from using $(CC) to $(CXX) for compilations
Diane Trout [Thu, 22 Sep 2005 00:46:01 +0000 (00:46 +0000)]
[project @ 14]
Fix save_path_base undefined error. Ticket:1
Diane Trout [Wed, 21 Sep 2005 23:26:49 +0000 (23:26 +0000)]
[project @ 13]
mussa015
Diane Trout [Wed, 21 Sep 2005 23:23:41 +0000 (23:23 +0000)]
[project @ 12]
mussa014 example
Diane Trout [Wed, 21 Sep 2005 23:22:49 +0000 (23:22 +0000)]
[project @ 11]
mussa014
Diane Trout [Wed, 21 Sep 2005 23:18:36 +0000 (23:18 +0000)]
[project @ 10]
mussa010
Diane Trout [Wed, 21 Sep 2005 23:17:51 +0000 (23:17 +0000)]
[project @ 9]
mussa009
Diane Trout [Wed, 21 Sep 2005 23:16:55 +0000 (23:16 +0000)]
[project @ 8]
mussa_008
Diane Trout [Wed, 21 Sep 2005 23:14:15 +0000 (23:14 +0000)]
[project @ 7]
mussa_pre.7
Diane Trout [Fri, 16 Sep 2005 23:08:29 +0000 (23:08 +0000)]
[project @ 6]
musssa_pre.5
Diane Trout [Fri, 16 Sep 2005 22:49:56 +0000 (22:49 +0000)]
[project @ 5]
mussa_pre.4.tar.gz
Diane Trout [Fri, 16 Sep 2005 22:45:17 +0000 (22:45 +0000)]
[project @ 4]
mussa_pI.3
Diane Trout [Fri, 16 Sep 2005 22:43:01 +0000 (22:43 +0000)]
[project @ 3]
checked in mussa_pI.2
Diane Trout [Fri, 16 Sep 2005 22:41:59 +0000 (22:41 +0000)]
[project @ 2]
Checked in mussa_pI.1