jajupmochi
6aac5b2981
1. add Sylvester Equation Methods for the generalized random walk kernel.
2. correct an error in the common walk kernel. DON NOT use the old one.
3. improve the method to construct fully-labeled direct product graphs, much faster for sparse graphs.
7 years ago
jajupmochi
efca08fb30
1. upgrate spkernel, now supporting symbolic and non-symbolic node labels, directed and undirected graphs, and allowing user-defined node kernels.
2. model_selection_precomputed can now save all results as human readable text.
3. modify pygraph.utils.utils.floydTransformation and pygraph.utils.graphdataset.get_dataset_attributes.
7 years ago
jajupmochi
bc93c0f80b
I am dead...
7 years ago
jajupmochi
ff2201d42b
Merge branch 'ljia' of https://github.com/jajupmochi/py-graph into ljia
7 years ago
jajupmochi
f684f51ad1
delete useless results.
7 years ago
jajupmochi
1ee3ee0cba
Merge branch 'ljia' of https://github.com/jajupmochi/py-graph into ljia
7 years ago
jajupmochi
503187dc3f
delete useless results files.
7 years ago
jajupmochi
609a7a0124
Merge branch 'ljia' of https://git.litislab.fr/bgauzere/py-graph into ljia
7 years ago
jajupmochi
b0f6470c35
Plus de choses sont changées...
7 years ago
jajupmochi
2db913b7eb
MOD extend marginalized kernel, remove tottering.
MOD modify tree pattern kernel, remove tottering.
MOD save output in function model_selection_precpmputed.
7 years ago
jajupmochi
c615605b08
MOD extend marginalized kernel, remove tottering.
MOD modify tree pattern kernel, remove tottering.
MOD save output in function model_selection_precpmputed.
7 years ago
jajupmochi
b683eb9e5e
remove unnecessary files.
7 years ago
jajupmochi
15b6a97053
modifié : README.md
modifié : notebooks/run_treepatternkernel.ipynb
modifié : notebooks/run_untilnwalkkernel.ipynb
7 years ago
jajupmochi
338f8ba326
modifié : README.md
modifié : notebooks/run_cyclicpatternkernel.ipynb
modifié : notebooks/run_marginalizedkernel_acyclic.ipynb
modifié : notebooks/run_pathkernel_acyclic.ipynb
modifié : notebooks/run_spkernel_acyclic.ipynb
modifié : notebooks/run_treeletkernel_acyclic.ipynb
modifié : notebooks/run_treepatternkernel.ipynb
modifié : notebooks/run_untildpathkernel_acyclic.ipynb
nouveau fichier : notebooks/run_untilnwalkkernel.ipynb
modifié : notebooks/run_weisfeilerLehmankernel_acyclic.ipynb
modifié : pygraph/kernels/treePatternKernel.py
modifié : pygraph/kernels/untildPathKernel.py
nouveau fichier : pygraph/kernels/untilnWalkKernel.py
nouveau fichier : pygraph/utils/model_selection_precomputed.py
modifié : pygraph/utils/utils.py
7 years ago
jajupmochi
0d82ebe6b8
copié : notebooks/run_weisfeilerLehmankernel_acyclic.ipynb -> .ipynb_checkpoints/run_weisfeilerLehmankernel_acyclic-checkpoint.ipynb
modifié : README.md
nouveau fichier : notebooks/run_cyclicpatternkernel.ipynb
modifié : notebooks/run_marginalizedkernel_acyclic.ipynb
modifié : notebooks/run_pathkernel_acyclic.ipynb
modifié : notebooks/run_spkernel_acyclic.ipynb
modifié : notebooks/run_treeletkernel_acyclic.ipynb
nouveau fichier : notebooks/run_treepatternkernel.ipynb
modifié : notebooks/run_weisfeilerLehmankernel_acyclic.ipynb
nouveau fichier : pygraph/kernels/cyclicPatternKernel.py
modifié : pygraph/kernels/deltaKernel.py
modifié : pygraph/kernels/pathKernel.py
modifié : pygraph/kernels/results.md
modifié : pygraph/kernels/spKernel.py
nouveau fichier : pygraph/kernels/treePatternKernel.py
modifié : pygraph/kernels/treeletKernel.py
modifié : pygraph/kernels/untildPathKernel.py
modifié : pygraph/kernels/weisfeilerLehmanKernel.py
modifié : pygraph/utils/graphfiles.py
modifié : pygraph/utils/utils.py
nouveau fichier : run_cyclic.py
7 years ago
jajupmochi
24eae65154
* ADD *path kernel up to depth d* and its result on dataset Asyclic.
* MOD treelet kernel, retrieve canonkeys of all graphs before calculate kernels, wildly speed it up.
7 years ago
jajupmochi
713edf2794
ADD comments to code of treelet kernel.
7 years ago
jajupmochi
a719ac6859
* ADD *treelet kernel* and its result on dataset Asyclic. - linlin
* MOD the way to calculate WL subtree kernel, correct its results. - linlin
* ADD *kernel_train_test* and *split_train_test* to wrap training and testing process. - linlin
* MOD readme.md file, add detailed results of each kernel. - linlin
7 years ago
jajupmochi
5d0d9e350b
* ADD *treelet kernel* and its result on dataset Asyclic. - linlin
* MOD the way to calculate WL subtree kernel, correct its results. - linlin
* ADD *kernel_train_test* and *split_train_test* to wrap training and testing process. - linlin
* MOD readme.md file, add detailed results of each kernel. - linlin
7 years ago
jajupmochi
460e846a8f
* ADD calculation of the time spend to acquire kernel matrices for each kernel. - linlin
* MOD floydTransformation function, calculate shortest paths taking into consideration user-defined edge weight. - linlin
* MOD implementation of nodes and edges attributes genericity for all kernels. - linlin
* ADD detailed results file results.md. - linlin
* MOD Weisfeiler-Lehman subtree kernel and the test code. - linlin
7 years ago
jajupmochi
6e7098004c
* MOD Implementation of attributes genericity for both gxl and ct functions
7 years ago
jajupmochi
1cac6d23af
* ADD Weisfeiler-Lehman subtree kernel and its result on dataset Asyclic.
7 years ago
jajupmochi
0184a2a06b
* ADD mean average path kernel and its result on dataset Asyclic. - linlin
* ADD delta kernel. - linlin
* MOD reconstruction the code of marginalized kernel. - linlin
7 years ago
jajupmochi
8ef59310f9
Merge branch 'ljia' of https://git.litislab.fr/bgauzere/py-graph into ljia
7 years ago
jajupmochi
5fa8e5aeea
initial
7 years ago
Benoit GAUZERE
4cd3654e90
Remove non pertinent files
7 years ago
jajupmochi
d8a96ce408
Merge branch 'master' of https://git.litislab.fr/bgauzere/py-graph
add pygraph/kernels/spkernel.py
modify pygraph/utils/util.py and pygraph/utils/graphfiles.py
7 years ago
jajupmochi
4ade070e23
commit
7 years ago
Benoit GAUZERE
66129a66df
Update to NetworkX 2.0
7 years ago
Benoit GAUZERE
24a8ec3ad3
Re structuration of project
7 years ago
Benoit GAUZERE
eda860455d
Re structuration of project
7 years ago
Benoit GAUZERE
d9dfc8ec68
Update of costfunctions hierarchy to distinguish cost function for ged and for cost matrix
7 years ago
Benoit GAUZERE
a3e75c82b7
This notebook compute ged approx using 2 methods and 2 solvers
7 years ago
Benoit GAUZERE
91cb5d2ee6
First init - may contains some errors...
7 years ago