Stephen McQuay
eadfa46e57
middle of refactor
2010-10-23 13:49:15 -06:00
Stephen McQuay
431f3a4e44
forgot the __init__.py file
2010-10-22 16:21:41 -06:00
Stephen McQuay
2db4169bfa
major refactoring
...
--HG--
rename : lib/baker/__init__.py => interp/baker/__init__.py
rename : lib/grid/DD.py => interp/grid/DD.py
rename : lib/grid/DDD.py => interp/grid/DDD.py
rename : lib/grid/__init__.py => interp/grid/__init__.py
rename : lib/grid/qhull.py => interp/grid/qhull.py
rename : lib/grid/simplex.py => interp/grid/simplex.py
rename : lib/grid/smcqdelaunay.py => interp/grid/smcqdelaunay.py
rename : lib/baker/tools.py => interp/tools.py
2010-10-22 16:10:58 -06:00
Stephen McQuay
98d7247497
minor: doc change
2010-10-22 09:21:04 -06:00
Stephen McQuay
91c8a661f4
working 2D gmsh parser
2010-10-22 09:20:59 -06:00
Stephen McQuay
e4988fd150
implementing a 2D mesh object for gmsh files
2010-10-06 12:15:09 -06:00
Stephen McQuay
b3680848d5
testing push
2010-09-04 16:52:16 -06:00
Stephen McQuay
73322f7554
testing push
2010-09-04 16:51:05 -06:00
Stephen McQuay
df35cb174b
made a get_error function that is generic and functional. I'm still getting unsavory numbers for the cubic, 2-D case on my single test case. I will have to try out more refined grids, albeit the quadratic is consistently improving results for my case.
2010-05-05 23:03:12 -06:00
Stephen McQuay
3a1c13bcac
implemented a function that will be used to provide a generic n-th order, nth-dimension error approximation function
2010-05-04 23:03:07 -06:00
c5adab295b
minor: but i'm on the verge of implementing the cubic error term for 3D
2010-04-30 16:23:29 -06:00
a1c88a186e
something about the changes I made (abs -> np.abs) made my stuff succeed more often
2010-04-30 11:56:01 -06:00
Stephen McQuay
2cbd92e15b
wrapping up a night. there isn't enough consistent improvement to merit using this method. i must have a bug somewhere.
2010-04-29 23:29:35 -06:00
3dcc10ea0e
working on the containing simplex problem. the get containing simplex function works, if the point is in the domain. I am trying to force the domain to contain the point (placing points along perimiter of the mesh). I am not doing that right, but have to run
...
--HG--
rename : bin/qhull-029.txt => data/qhull-029.txt
rename : bin/qhull-029.txt.gv => data/qhull-029.txt.gv
2010-04-24 18:26:44 -06:00
cfcd1e15a2
working on the guarantee simplex routine. I have a test case to try it out against, and need to look at visiting neighbors of faces adjacent to points.
2010-04-23 16:29:31 -06:00
0a388ff1b5
trying to guarantee a containing simplex
2010-04-23 09:58:30 -06:00
79880acad4
fixed discrepancy with Baker's eq 17.
2010-04-09 15:03:16 -06:00
3efe335563
moved the quadratic error calculation out of the baker method, and added a cubic version too
2010-04-02 11:06:17 -06:00
700ccc8c25
created a test for tesing the cubic interpolation
2010-03-20 20:09:46 -06:00
1fc988428d
putting together a simple test case so that I can test my quad/cubic interpolator
2010-03-20 17:10:02 -06:00
993c901cec
properly catching the LinAlgError. also, prepping to change the run_baker method (more generic)
2010-03-20 11:34:24 -06:00
bd8f24d64f
working on implementing the cubic interpolation. need to split that routine out of the run_baker method
2010-03-19 21:27:51 -06:00
f089848a1b
i think part of my problem (why i'm not always getting win), is that i do not guarantee that the simplex contains the point (especially with the random clounds; i think that the connectivity-based lookup is doing that fine
2010-03-18 23:50:40 -06:00
9836fdcbfe
added files. still not consistently getting better numbers
2010-03-18 23:32:24 -06:00
b33159f8a9
working on getting baker to work in 3D. the code runs, but the numbers are odd. I suspect the rectangular grid, and am going to try a random cloud of points.
2010-03-18 23:18:59 -06:00
a2d7b3f063
updated library files so that i don't have to edit the __init__.py files
...
--HG--
rename : lib/baker/__init__.py => lib/baker/baker.py
rename : lib/grid/__init__.py => lib/grid/grid.py
2010-03-08 14:30:22 -07:00
b9ea6a3ac2
modified code to use Distribute (python's distribution framework)
...
one should simply run `python setup.py install` to install these librarires
2010-03-08 13:31:51 -07:00
c5d8fff883
moved some files around to conform to python best practicies
...
--HG--
rename : lib/baker.py => lib/baker/__init__.py
rename : lib/tools.py => lib/baker/tools.py
rename : lib/grid.py => lib/grid/__init__.py
rename : lib/smcqdelaunay.py => lib/grid/smcqdelaunay.py
2010-03-08 13:05:42 -07:00
9a1b8d14b2
finally added proper (in my opinon) exception handling to the grid's run_baker method. it should properly autocorrect, and switch the extra point lookup scheme to connectivity-based if solving the system of equations yields a singular solution
2010-03-05 08:58:07 -07:00
a25e6d03d1
partially implemented using new data structure in point lookup
...
I spoke with Dr. Gorrell about first trying to call the nearest-neighbor routine, releasing the trapped exception where the matrix ends up being singular, and trapping it in the grid object's run_baker, where I could then call the connectivity-based lookup routine. Those are my next steps, as well as starting to think about pushing the 3D front ahead this week.
2010-02-26 14:44:29 -07:00
2792329eaa
cleaned up the method (removed print statements) i'm ready to now add this method as an option to the mesh object
2010-02-21 20:52:51 -07:00
a3d72ed200
wrote the method that returns grid objects based on connectivity
...
I'm going to hurry and clean up this method, but for now it is verbose, and uses sets. i'm going to just check it in for future reference
2010-02-21 20:42:32 -07:00
0ef6897aee
made a simplex class, going to move this into the face class to simplify
2010-02-21 19:02:45 -07:00
6c4c6d4cfe
moved teh grid code into the grid class
...
by next friday i need to have implemented a way to select between normal kdtree point lookups, and nearest-neighbor lookups. Hopefully i'll have that done tonight.
2010-02-20 13:18:24 -07:00
2cf9a0b574
mostly fixed a bug in my qdelaunay parsing. implemented a blender viewer ... mostly
...
the qdelaunay app doesn't alway display all faces for a point. lame. I worked around it by creating the point-> face relationships manually.
There is a max recursion something happeneing when I try to save so pickle. look into this.
I spent some time writing a visualizer for blender. 10 lines of code. simple stuff.
next steps would be to move the KDTree into the mesh object. GOOD LUCK (me)!!
--HG--
rename : bin/grid_driver.py => bin/grid_random.py
rename : bin/grid_driver.py => bin/grid_regular.py
2010-02-14 20:56:41 -07:00
9ff48f4c12
minor: added point to facet information to the grid object
...
I really think that I need to create my own point object, and hang the facet information directly off of that. right now there are simply parallel arrays for q, numpy.array(x,y), and {point_index: [adjacent faces]} ...
work to be done!
2010-02-10 00:20:06 -07:00
9064663600
added face information to the grid. i need to parse and add point->face information ... somehow ...
2010-02-07 20:23:19 -07:00
50daeb5d74
getting stuff from qdelaunay
...
mostly spent time trying to figure out subprocess (which i though i already understood). I am concerned for what will happen with large datasets.
next steps include:
1) design data struture
2) implement slurping of stdout into this structure
3) drink coke
--HG--
rename : lib/smcqhull.py => lib/smcqdelaunay.py
2010-02-07 00:01:02 -07:00
f9973e7580
minor, updated the driver to handle linear case
2010-01-31 21:15:47 -07:00
0f4daa9a9c
fixed a little bug and getting good results again. still will have problems if you try to simply get a linear interpolation
2010-01-31 21:10:37 -07:00
2113abb8dc
minor: updated the .setup file
2010-01-31 20:51:25 -07:00
80720c45fe
MAJOR: updated the baker method. it's more generic, and should allow me to 3D-ifiy it more simply. a ton of other things
2010-01-31 20:47:03 -07:00
98b13fb8c5
merged the drivers, and am investigating the qdelauney manpage ... should've done this earlier :P
...
--HG--
rename : bin/driver-random.py => bin/driver.py
rename : test/utest.py => test/qhull.test.py
2010-01-30 12:15:00 -07:00
f4b2c95cf5
working on adding 3D baker. also starting to massage together a good baker method test (2D), and added a stub file for 3D testing
2010-01-29 11:56:52 -07:00
961b3b7b26
init
2009-12-27 10:48:27 -07:00
Stephen McQuay
Stephen Mardson McQuay
smcquay@cfdviz2