deb3f656cd
I have added an MIT license to the code, including all authors who have contributed to Master at this point in time. Each of you should approve this commit and then we can merge it into master.
2013-08-06 18:26:49 -07:00
c3476321a8
Removed EMforward code for merging into master. (This is still being developed in branch: eldadsWork)
2013-08-06 18:07:21 -07:00
7e5ea835e3
Removed eldadCode directory as it is now integrated. Added GaussNewton to the main directory.
2013-08-06 17:56:21 -07:00
3434ea83ba
Merge branch 'master' of https://bitbucket.org/rcockett/simpeg into LOM
2013-08-06 17:52:09 -07:00
43d2512aea
Updated the test_operators code to work for LOM. need to be careful in which error metric you consider (because some only make sense under integration)
...
i.e. the l-2 norm rather than the infinity norm, and you need to multiply by the integration length/area/volume it you are looking at the l-2 norm.
2013-08-06 17:18:04 -07:00
262aacce85
Update tests with new calls. Note that there seem to be problems with rotateLOM on some operator tests.
2013-08-06 15:09:01 -07:00
a510755926
Moved projectFace/EdgeVector to the base mesh class, along with some default normal and tangents.
2013-08-06 15:07:32 -07:00
f01b4370b9
Added example of plotting the grid for a Lom
2013-08-05 17:28:55 -07:00
4b545fb1b7
Cleaned up projection code and put it in LOM.
2013-08-05 17:16:01 -07:00
83fe9df743
Ha. Not a bug, the test was wrong. Must project the components of the field.
2013-08-05 16:48:05 -07:00
3d936e07a1
Edge Inner products in 2D NOTE: still has bugs.
2013-08-05 12:16:55 -07:00
e073eaeb8b
Innerproduct work. Testing, visualization, and 2D - NOTE there are bugs in anything but a uniform LOM
...
Not very helpful yet!
2013-08-05 11:51:03 -07:00
e2e38074fc
Order Test must incorporate different meshes. Current test fails, must support 2D mass matrices.
2013-08-04 07:56:25 -07:00
849a914738
LOM grid ding and visualization code.
2013-08-04 07:54:33 -07:00
26b334585f
EdgeInnerProducts now working for LOM
...
Fixed DiffOps bug in CURL
Fixed bug in OrderTest --> must remember to do cumSum, and not just supply dx to ndgrid
test_massMatrices now tests uniformLOM
2013-08-03 16:39:34 -07:00
cc66eaf9fd
Testing of LOM (comparing against Tensor Mesh)
2013-08-03 16:08:54 -07:00
1acb1a4b84
Ensure you return None if dimension of TensorMesh is less than the requested grid.
2013-08-03 16:08:04 -07:00
6d5190b2f2
GridFunctions!
2013-08-03 16:07:27 -07:00
9ca9c20731
Modified innerProducts so they have defaults, and you have to explicitly ask for the projection matrices.
2013-08-03 13:57:00 -07:00
333b484cfb
Merge branch 'eldadswork' of https://bitbucket.org/rcockett/simpeg into LOM
2013-08-03 13:55:04 -07:00
306e31ef04
Added warning to the faceInfo utils ---> this does not compute concave areas correctly.
...
This is an issue, and should be fixed in future versions.
2013-08-03 13:53:08 -07:00
caaa7dcb41
made the inner products easy to use to get the derivatives with respect to conductivity/mu
2013-08-03 10:23:35 -07:00
24773c4f60
my first attempt at a wrapper to overcome some of numpy issues
2013-08-03 02:07:15 -07:00
f41463b241
need more work but its closer
2013-08-02 01:51:46 -07:00
ec4358d872
corrected bugs
2013-08-02 01:51:10 -07:00
dcae030d81
1. Added output for the edgeInnerProduct so we can use it in the derivative phase
...
2. The getMisfit may be working but needs a little bit more work
We need to go through our code to make sure that all vectors that can be multiplied by a matrix are set to
vec = mkvc(vec,2)
2013-08-02 00:18:46 -07:00
d3c4f06876
Merge branch 'master' of https://bitbucket.org/rcockett/simpeg into condModels
2013-08-01 23:15:22 -07:00
a3ecaf7996
Added an __str__ method to the mesh class. Will allow for basic mesh info to be viewed using 'print mesh'
2013-08-01 22:54:52 -07:00
444feca101
initial forward EM
2013-08-01 15:05:12 -07:00
4887f4b130
Minor changes to utile so that it works with matrices.
2013-08-01 08:23:23 -07:00
5af4a5f5c2
3D volume tweak now takes the average of two different ways to divide into tetrahedras
2013-07-31 18:34:03 -07:00
ea223d073c
Face Normals.
...
Note that there are some serious differences in how these are stored based in 3D vs 2D, but that is the nature of the beast.
:)
2013-07-31 18:07:52 -07:00
34e380ccb9
edge and tangent calculating, along with testing functions.
2013-07-31 15:07:14 -07:00
b07256bbc5
Merge branch 'eldadswork' of https://bitbucket.org/rcockett/simpeg into LOM
...
Conflicts:
SimPEG/utils.py
2013-07-31 11:30:57 -07:00
451e4260ce
Removed x0 from input for LOM
2013-07-31 11:22:21 -07:00
47c0c7603d
Area and indexCube bug fix
2013-07-30 23:43:13 -07:00
7e169bbb79
Cleaned utils code, merged in subArray.
2013-07-30 22:56:36 -07:00
68e44cca5a
Deleted massMatrices.py
...
This code is not working and is repeated in InnerProducts.py
2013-07-30 22:52:37 -07:00
4660b177a1
Names in tests.
2013-07-30 22:48:27 -07:00
9573c6c230
Tested face and edge inner products for anisotropic, isotropic and tensor sigma.
2013-07-30 22:40:10 -07:00
c259a6651a
Integrated getEdge/FaceInnerProduct into the tensor mesh class.
2013-07-30 22:12:53 -07:00
d8f646c7d9
Updates to OrderTest so that it runs random meshes.
2013-07-30 21:49:23 -07:00
ead836c341
Fixed 2D volume calculation.
2013-07-29 23:02:25 -07:00
1fca1b894e
fixed interpmat need more testing
2013-07-29 12:47:41 -07:00
f7713656ef
Added interpolation matrix. Not sure its working since my canopy is playing tricks
2013-07-27 17:36:00 -07:00
05ddf93d64
Added getFaceInnerProduct and getInnerProduct that puts the edge and face mass matrices together
2013-07-27 16:11:24 -07:00
079acb1153
volume in 2D by calculating faceInfo (in utils for now.)
2013-07-26 16:38:12 -07:00
fe547476c9
IndexCube now in Utils
2013-07-26 15:56:21 -07:00
b4bb1024e4
Merge branch 'eldadswork' of https://bitbucket.org/rcockett/simpeg into LOM
...
Conflicts:
SimPEG/utils.py
2013-07-26 15:33:35 -07:00
5c83095781
Minor changes and comments.
2013-07-26 14:30:04 -07:00
Rowan Cockett