# graphkit-learn
[](https://app.travis-ci.com/jajupmochi/graphkit-learn)
[](https://ci.appveyor.com/project/jajupmochi/graphkit-learn)
[](https://codecov.io/gh/jajupmochi/graphkit-learn)
[](https://graphkit-learn.readthedocs.io/en/master/?badge=master)
[](https://badge.fury.io/py/graphkit-learn)
[](https://gitter.im/graphkit-learn/graphkit-learn?utm_source=badge&utm_medium=badge&utm_campaign=pr-badge&utm_content=badge)
A Python package for graph kernels, graph edit distances and graph pre-image problem.
## Requirements
* python>=3.6
* numpy>=1.16.2
* scipy>=1.1.0
* matplotlib>=3.1.0
* networkx>=2.2
* scikit-learn>=0.20.0
* tabulate>=0.8.2
* tqdm>=4.26.0
* control>=0.8.2 (for generalized random walk kernels only)
* slycot>=0.3.3 (for generalized random walk kernels only, which requires a fortran compiler (e.g., `gfortran`) and BLAS/LAPACK (e.g. `liblapack-dev`))
## How to use?
### Install the library
* Install stable version from PyPI (may not be up-to-date):
```
$ pip install graphkit-learn
```
* Install latest version from GitHub:
```
$ git clone https://github.com/jajupmochi/graphkit-learn.git
$ cd graphkit-learn/
$ python setup.py install
```
### Run the test
A series of [tests](https://github.com/jajupmochi/graphkit-learn/tree/master/gklearn/tests) can be run to check if the library works correctly:
```
$ pip install -U pip pytest codecov coverage pytest-cov
$ pytest -v --cov-config=.coveragerc --cov-report term --cov=gklearn gklearn/tests/
```
### Check examples
A series of demos of using the library can be found on [Google Colab](https://drive.google.com/drive/folders/1r2gtPuFzIys2_MZw1wXqE2w3oCoVoQUG?usp=sharing) and in the [`example`](https://github.com/jajupmochi/graphkit-learn/tree/master/gklearn/examples) folder.
### Other demos
Check [`notebooks`](https://github.com/jajupmochi/graphkit-learn/tree/master/notebooks) directory for more demos:
* [`notebooks`](https://github.com/jajupmochi/graphkit-learn/tree/master/notebooks) directory includes test codes of graph kernels based on linear patterns;
* [`notebooks/tests`](https://github.com/jajupmochi/graphkit-learn/tree/master/notebooks/tests) directory includes codes that test some libraries and functions;
* [`notebooks/utils`](https://github.com/jajupmochi/graphkit-learn/tree/master/notebooks/utils) directory includes some useful tools, such as a Gram matrix checker and a function to get properties of datasets;
* [`notebooks/else`](https://github.com/jajupmochi/graphkit-learn/tree/master/notebooks/else) directory includes other codes that we used for experiments.
### Documentation
The docs of the library can be found [here](https://graphkit-learn.readthedocs.io/en/master/?badge=master).
## Main contents
### 1 List of graph kernels
* Based on walks
* [The common walk kernel](https://github.com/jajupmochi/graphkit-learn/tree/master/gklearn/kernels/common_walk.py) [1]
* Exponential
* Geometric
* [The marginalized kernel](https://github.com/jajupmochi/graphkit-learn/tree/master/gklearn/kernels/marginalized.py)
* With tottering [2]
* Without tottering [7]
* [The generalized random walk kernel](https://github.com/jajupmochi/graphkit-learn/tree/master/gklearn/kernels/random_walk.py) [3]
* [Sylvester equation](https://github.com/jajupmochi/graphkit-learn/tree/master/gklearn/kernels/sylvester_equation.py)
* Conjugate gradient
* Fixed-point iterations
* [Spectral decomposition](https://github.com/jajupmochi/graphkit-learn/tree/master/gklearn/kernels/spectral_decomposition.py)
* Based on paths
* [The shortest path kernel](https://github.com/jajupmochi/graphkit-learn/tree/master/gklearn/kernels/shortest_path.py) [4]
* [The structural shortest path kernel](https://github.com/jajupmochi/graphkit-learn/tree/master/gklearn/kernels/structural_sp.py) [5]
* [The path kernel up to length h](https://github.com/jajupmochi/graphkit-learn/tree/master/gklearn/kernels/path_up_to_h.py) [6]
* The Tanimoto kernel
* The MinMax kernel
* Non-linear kernels
* [The treelet kernel](https://github.com/jajupmochi/graphkit-learn/tree/master/gklearn/kernels/treelet.py) [10]
* [Weisfeiler-Lehman kernel](https://github.com/jajupmochi/graphkit-learn/tree/master/gklearn/kernels/weisfeiler_lehman.py) [11]
* [Subtree](https://github.com/jajupmochi/graphkit-learn/tree/master/gklearn/kernels/weisfeiler_lehman.py#L479)
A demo of computing graph kernels can be found on [Google Colab](https://colab.research.google.com/drive/17Q2QCl9CAtDweGF8LiWnWoN2laeJqT0u?usp=sharing) and in the [`examples`](https://github.com/jajupmochi/graphkit-learn/blob/master/gklearn/examples/compute_graph_kernel.py) folder.
### 2 Graph Edit Distances
### 3 Graph preimage methods
A demo of generating graph preimages can be found on [Google Colab](https://colab.research.google.com/drive/1PIDvHOcmiLEQ5Np3bgBDdu0kLOquOMQK?usp=sharing) and in the [`examples`](https://github.com/jajupmochi/graphkit-learn/blob/master/gklearn/examples/median_preimege_generator.py) folder.
### 4 Interface to `GEDLIB`
[`GEDLIB`](https://github.com/dbblumenthal/gedlib) is an easily extensible C++ library for (suboptimally) computing the graph edit distance between attributed graphs. [A Python interface](https://github.com/jajupmochi/graphkit-learn/tree/master/gklearn/gedlib) for `GEDLIB` is integrated in this library, based on [`gedlibpy`](https://github.com/Ryurin/gedlibpy) library.
### 5 Computation optimization methods
* Python’s `multiprocessing.Pool` module is applied to perform **parallelization** on the computations of all kernels as well as the model selection.
* **The Fast Computation of Shortest Path Kernel (FCSP) method** [8] is implemented in *the random walk kernel*, *the shortest path kernel*, as well as *the structural shortest path kernel* where FCSP is applied on both vertex and edge kernels.
* **The trie data structure** [9] is employed in *the path kernel up to length h* to store paths in graphs.
## Issues
* This library uses `multiprocessing.Pool.imap_unordered` function to do the parallelization, which may not be able to run correctly under Windows system. For now, Windows users may need to comment the parallel codes and uncomment the codes below them which run serially. We will consider adding a parameter to control serial or parallel computations as needed.
* Some modules (such as `Numpy`, `Scipy`, `sklearn`) apply [`OpenBLAS`](https://www.openblas.net/) to perform parallel computation by default, which causes conflicts with other parallelization modules such as `multiprossing.Pool`, highly increasing the computing time. By setting its thread to 1, `OpenBLAS` is forced to use a single thread/CPU, thus avoids the conflicts. For now, this procedure has to be done manually. Under Linux, type this command in terminal before running the code:
```
$ export OPENBLAS_NUM_THREADS=1
```
Or add `export OPENBLAS_NUM_THREADS=1` at the end of your `~/.bashrc` file, then run
```
$ source ~/.bashrc
```
to make this effective permanently.
## Results
Check this paper for detailed description of graph kernels and experimental results:
Linlin Jia, Benoit Gaüzère, and Paul Honeine. Graph Kernels Based on Linear Patterns: Theoretical and Experimental Comparisons. working paper or preprint, March 2019. URL https://hal-normandie-univ.archives-ouvertes.fr/hal-02053946.
A comparison of performances of graph kernels on benchmark datasets can be found [here](https://gra
Python 图机器学习算法库.rar （1976个子文件） 
libdoublefann.so.2.2.0 124KB libfloatfann.so.2.2.0 124KB libfann.so.2.2.0 124KB libfixedfann.so.2.2.0 81KB libdoublefann.so.2 22B libfloatfann.so.2 21B libfixedfann.so.2 21B libfann.so.2 16B make.bat 821B make.bat 791B svmtrain.c 12KB svmpredict.c 10KB svm-train.c 9KB svm-scale.c 8KB svm_model_matlab.c 8KB interface.c 6KB svm-predict.c 5KB libsvmread.c 4KB libsvmwrite.c 2KB main.c 398B COPYRIGHT 1KB .coveragerc 51B gedlibpy.cpp 1.08MB svm.cpp 64KB svm-toy.cpp 11KB callbacks.cpp 10KB svm-toy.cpp 10KB cpp_code.cpp 5KB tribenzo[a,e,i]pyrene.ct 2KB dibenzo[b,k]chrysene.ct 1KB dibenzo[a,j]naphthacene.ct 1KB benzo[c]pentaphene.ct 1KB naphtho[1,2-b]triphenylene.ct 1KB anthra[1,2-a]anthracene.ct 1KB dibenzo[a,l]naphthacene.ct 1KB coronene.ct 1KB dibenzo[c,i]pyrene.ct 1KB dibenzo[a,h]pyrene.ct 1KB 6,12-dimethylanthanthrene.ct 1KB dibenzo[a,e]pyrene.ct 1KB naphtho(2,3-a)pyrene.ct 1KB dibenzo[e,l]pyrene.ct 1KB dibenzo[a,l]pyrene.ct 1KB 6-methylanthanthrene.ct 1KB anthanthrene.ct 1KB benzo[g,h,i]perylene.ct 1KB pentaphene.ct 1KB dibenz[a,c]anthracene.ct 1KB dibenz[a,h]anthracene.ct 1KB 1,4-dimethylbenzo[a]pyrene.ct 1KB 4,5-dimethylbenzo[a]pyrene.ct 1KB 3,6-dimethylbenzo[a]pyrene.ct 1KB dibenzo[c,g]phenanthrene.ct 1KB pentacene.ct 1KB benzo[g]chrysene.ct 1KB benzo[a]naphthacene.ct 1KB picene.ct 1KB 3,12-dimethylbenzo[a]pyrene.ct 1KB 7,10-dimethylbenzo[a]pyrene.ct 1KB benzo[c]chrysene.ct 1KB benzo[b]chrysene.ct 1KB dibenz[a,j]anthracene.ct 1KB 2,3-dimethylbenzo[a]pyrene.ct 1KB 1,6-dimethylbenzo[a]pyrene.ct 1KB 1,3-dimethylbenzo[a]pyrene.ct 1KB dibenzo[b,g]phenanthrene.ct 1KB 1,2-dimethylbenzo[a]pyrene.ct 1KB 11-methylbenzo[a]pyrene.ct 1KB 4-methylbenzo[a]pyrene.ct 1KB 5-methylbenzo[a]pyrene.ct 1KB 12-methylbenzo[a]pyrene.ct 1KB 3-methylbenzo[a]pyrene.ct 1KB 7-methylbenzo[a]pyrene.ct 1KB 2-methylbenzo[a]pyrene.ct 1KB 1-methylbenzo[a]pyrene.ct 1KB 8-methylbenzo[a]pyrene.ct 1KB 10-methylbenzo[a]pyrene.ct 1KB 6-methylbenzo[a]pyrene.ct 1KB 6,8,12-trimethylbenz[a]anthracene.ct 1KB benzo[e]pyrene.ct 1KB perylene.ct 1KB benzo[a]pyrene.ct 1KB 7,12-dimethylbenz[a]anthracene.ct 1KB 1,12-dimethylbenz[a]anthracene.ct 1KB 5,8-dimethylbenzo[c]phenanthrene.ct 1KB 6,12-dimethylbenz[a]anthracene.ct 1KB 6,8-dimethylbenz[a]anthracene.ct 1KB 4-methylchrysene.ct 1KB 4-methylbenz[a]anthracene.ct 1KB 3-methylbenz[a]anthracene.ct 1KB 1-methylbenz[a]anthracene.ct 1KB 4-methylbenzo[c]phenanthrene.ct 1KB 7-methylbenz[a]anthracene.ct 1KB 3-methylbenzo[c]phenanthrene.ct 1KB 6-methylbenz[a]anthracene.ct 1KB 1-methylchrysene.ct 1KB 2-methylbenz[a]anthracene.ct 1KB 1-methylbenzo[c]phenanthrene.ct 1KB 11-methylbenz[a]anthracene.ct 1KB 5-methylchrysene.ct 1KB共 1976 条
- 1
- 2
- 3
- 4
- 5
- 6
- 20
资源评论
