Commit graph

2618 commits

Author SHA1 Message Date
Zhenwen Dai
8a4a4e56a9 fix pickle for ssgplvm and bgplvm with mpi 2014-06-26 14:39:02 +01:00
Zhenwen Dai
1512590a8c remove dependence of scikits.cuda from rbf kernel 2014-06-26 14:12:19 +01:00
Zhenwen Dai
92fecbb4ef remove dependence of cublas from rbf kernel 2014-06-26 14:01:58 +01:00
Zhenwen Dai
c3482e7a94 fix the gpu initialization for multiple cards 2014-06-26 13:46:47 +01:00
Zhenwen Dai
bdf478956d adapt gpu initialization multiple gpu cards 2014-06-26 12:17:45 +01:00
Zhenwen Dai
cf33808673 fix the SSGPLVM with MPI 2014-06-25 13:57:25 +01:00
Zhenwen Dai
08ed72b2f2 fix the speed problem of the tie framework 2014-06-24 18:01:16 +01:00
Zhenwen Dai
525d9329ff a little optimization of gpu code 2014-06-24 17:28:01 +01:00
Zhenwen Dai
9c6bfae0b9 RBF for SSGPLVM gpu implemented 2014-06-24 15:10:53 +01:00
Zhenwen Dai
3f36a245d1 fix the problem of multiple ties on the same param array object 2014-06-24 12:49:04 +01:00
Zhenwen Dai
567612b3a9 tie framework works roughly 2014-06-23 19:39:35 +01:00
Zhenwen Dai
fa6b5343cc merge ties branch into psi2 2014-06-23 14:37:11 +01:00
Zhenwen Dai
36d53f815c merge ties branch into psi2 2014-06-23 14:36:28 +01:00
Zhenwen Dai
cf2bf3bbe2 var_dtc_parallel make YY.T speed up 2014-06-23 14:23:43 +01:00
Zhenwen Dai
7a35c9a6ad fix pickle for RBF GPU kernel 2014-06-23 12:02:36 +01:00
Zhenwen Dai
59172435e2 fix the remaining problem of cache.py 2014-06-20 21:51:51 +01:00
Zhenwen Dai
ca1edecce4 rbf gpu usable 2014-06-20 18:02:35 +01:00
Zhenwen Dai
e486f3fd99 fix import issue on no-gpu machine 2014-06-20 15:03:44 +01:00
Zhenwen Dai
155511b761 rbf kernel gpu implementation ready 2014-06-20 15:02:00 +01:00
Zhenwen Dai
e0412ebf54 rbf gpu psicomp pass gradcheck 2014-06-20 14:09:26 +01:00
Zhenwen Dai
db9b9bc857 rbf kernel gpu implementation in progress 2014-06-20 11:56:21 +01:00
Zhenwen Dai
0b75aa8b0f add truncated linear kernel 2014-06-16 10:14:41 +01:00
Zhenwen Dai
c2a1ea9cbd rename sslinear_psi_comp.py 2014-06-09 14:24:20 +01:00
Zhenwen Dai
695bad63c5 fix Linear kernel with SSGPLVM 2014-06-09 14:21:56 +01:00
Zhenwen Dai
6119c33078 Support non-symmetric dL_dKmm for stationary kernel 2014-06-03 11:34:32 +01:00
James Hensman
8eaa0bbf8a basic vardtc working 2014-05-30 16:45:51 +01:00
James Hensman
ac2d28e2fd reverting the fixing behaviour.
two reasons: 1) the new behaviour is confusing for new users. Either
something is fixed, or it's not. 2) the fixing didn't work! things that
should have been fixed were passed to the optimizer for optimization.

If we really want to save keystrokes, consider this:

m.foo.fix()
m.foo.unfix()
m.foo.constrain_positive()

is the same as

m.foo.fix()
m.foo.constrain_positive()

but the latter throws a warning.
2014-05-29 09:55:49 +01:00
Zhenwen Dai
b945c20004 restructure rbf kernel 2014-05-21 16:42:35 +01:00
Zhenwen Dai
04ab93a961 BayersianGPLVM mpi support 2014-05-21 12:44:24 +01:00
Zhenwen Dai
52c0be1848 merge devel branch in 2014-05-21 10:38:34 +01:00
Zhenwen Dai
a2203179f6 [psi2] implement RBF cpu 2014-05-21 10:34:51 +01:00
Max Zwiessele
b520eb212c [fixing] fixing now saves the old constraint 2014-05-20 16:39:52 +01:00
Max Zwiessele
c507cfe4ab [index operations] added lookup for properties for a given index as dict <properties, subindex> for given index 2014-05-20 16:06:56 +01:00
Max Zwiessele
ef256223d1 [mrd] more control for init, some missing data adjustements, init greatly improved 2014-05-20 14:49:20 +01:00
Max Zwiessele
995de897b3 [core updates] first try to switch updates off and on. Use m.updates = False to switch updates off, and vice-versa 2014-05-20 14:48:53 +01:00
Max Zwiessele
1ebcee967a [psi-stats] add kernel was missing a psi zero call 2014-05-20 14:47:57 +01:00
Max Zwiessele
76718a8fcc [stationary] input_sensitiviy is now 1/(l**2) 2014-05-20 14:47:28 +01:00
Max Zwiessele
533e5fb744 [datasets] delete packed data in hapmap dataset 2014-05-20 14:46:29 +01:00
Max Zwiessele
99699e9e02 [vardtc missing data] can handle non broadcastable selections 2014-05-20 14:46:03 +01:00
Max Zwiessele
8c8d06c8ae [pca] colors now as iterator 2014-05-20 14:44:59 +01:00
Max Zwiessele
dfa23b77c5 [posteriot] adjusted for more then one covariance per output 2014-05-20 14:43:58 +01:00
Max Zwiessele
dafad62363 [lmv_dimselect] we need to keep a pointer to the lvm_dimselect object, as the updates are weak references: dim_select = ... 2014-05-20 14:43:18 +01:00
Zhenwen Dai
001db6b089 [mpi] enable checkgrad 2014-05-18 10:05:35 +01:00
Zhenwen Dai
e7177b6d37 [mpi] fix the bug of mpi 2014-05-17 22:18:00 +01:00
Max Zwiessele
01d6b91f90 [mrd] missing data implemented, and plotting better 2014-05-16 15:12:19 +01:00
Zhenwen Dai
e6d07ad5ac fix pickle 2014-05-16 11:23:44 +01:00
Max Zwiessele
94c84a23a3 [mrd] plotting, init, inference etc. 2014-05-16 11:21:08 +01:00
Max Zwiessele
efc1f4413c [ploting] dim reduction 2014-05-16 11:20:50 +01:00
Zhenwen Dai
17b6b94db3 Logistic transformation numerical robustness 2014-05-16 10:48:18 +01:00
Zhenwen Dai
c568bad4fb bug fix for mpi SSGPLVM 2014-05-16 10:40:45 +01:00