km_config

KO | EN

gitlite — search

#hacktoberfest #react #ai #rust #deep-learning #macos #nodejs #java #windows #awesome-list #mcp #ios #openai #open-source

(★ 13)

An algorithm for finding multiple core-periphery pairs in networks.

km_config Latest Version Download

Download Latest Version (.zip)

// repository documentation

Was this content helpful?

★ 0 (0 ratings)

# This project has been renamed and moved to https://github.com/skojaku/core-periphery-detection

# km_config
C++, Matlab and Python code for the KM--config algorithm.

Please cite  
  Kojaku, S. and Masuda, N. "Core-periphery structure requires somthing else in the network". New Journal of Physics, 20, (2017).

Table of Contents
=================
* [C\+\+](#c)
* [Matlab](#matlab)
* [Python](#python)
* [Files](#files)
* [Requirement](#requirement)

## C++

### Compile:

To compile, type
        
  ```bash
  make cpp
  ```
       
  This will produce an executable file "km_config" in the src/cpp directory.
 
 
### Usage:

``` bash
  km_config [input-file] [output-file] [options]
  ```

km_config seeks multiple core-periphery pairs in the network given by [input-file] and saves the detected core-periphery pairs in [output-file].
  
**[input-file]**  
 * The file should contain a list of edges (space-separated).  
 * The first and second columns represent the IDs of the two nodes forming an edge.
 * The third column represents the weight of the edge between the two nodes. 
 * If the third column is not provided, then the weight is set to 1.  
 * The node's ID is assumed to start from 1. You can change this by [options], e.g., -i 0.
  
**[output_file]**  
 * The first column represents the node's ID.
 * The second column represents the index of the core-periphery pair to which each node belongs.
 * The third column indicates whether each node is a core node (= 1) or a peripheral node (= 0).
 * The fourth column indicates whether each node belongs to a significant core-periphery pair (= 1) or not (= 0).
  
  
**[options]**  
* -r R  
 Run the KM-config algorithm R times. (Default: 10)  
* -a ALPHA  
  Set the significance level before the Šidák correction to ALPHA. (Default: 1). If this option is not set, the statistical test is not carried out.
* -l N  
Set the number of randomised networks to N. (Default: 500)
* -i I
The node's ID starts from I. (Default: 1)
* -d "D"  
Change the delimiter for [input-file] and [output-file] to D. (Default: space)

### Example (src/cpp/example.sh):
    
```bash
#To find core-periphery pairs, type
./km_config example_edge_list.txt result.txt
    
#To find significant core-periphery pairs at a significance level of 0.05, type
./km_config example_edge_list.txt result.txt -a 0.05 
```

## Matlab 
      
### Compile:

To compile, type
        
  ```bash
  make matlab
  ```
    
  This will produce a mex file "km_config_mex.mexa64" in the src/matlab/ directory. 
  Copy "matlab/km_config_mex.mexa64" and "src/matlab/km_config.m" to your working directory.

You may have the following message:
  
  ```bash
  Warning: You are using gcc version '5.4.0'. The version of gcc is not supported. 
  The version currently supported with MEX is '4.9.x'. 
  ```
  
  This means you are required to change the version of g++ compiler. 
  To remedy this, modify a line in ''./src/matlab/makefile'' as follows: 
  
  ```bash
  MEXCOMPILER := g++-(the version compatible with your mex compiler, e.g., g++-4.9) 
  ```

If you get erros like this: 
  ```bash
  Invalid MEX file '~/program/git/km_config/src/matlab/km_config_mex.mexa64': /usr/local/MATLAB/R2016b/bin/glnxa64/../../sys/os/glnxa64/libstdc++.so.6: version `GLIBCXX_3.4.21' not found (required by ~/program/git/km_config/src/matlab/km_config_mex.mexa64) 
  ```

then open `src/lib/km_config.h` and comment out lines 51--57, i.e.,  
  ```cpp
  %std::mt19937_64 init_random_number_generator(){
  %	mt19937_64 mtrnd;
  %    	random_device r;
  %    	seed_seq seed{ r(), r(), r(), r(), r(), r(), r(), r() };
  %    	mtrnd.seed(seed);
  %	return mtrnd;
  %}
  ```
  and remove lines 63 and 82 to make the following function callable:

```cpp
  std::mt19937_64 init_random_number_generator(){
      int seeds[624];
      size_t size = 624*4; //Declare size of data
      std::ifstream urandom("/dev/urandom", std::ios::in | std::ios::binary); //Open stream
      if (urandom) //Check if stream is open
      {
          urandom.read(reinterpret_cast<char*>(seeds), size); //Read from urandom
          urandom.close(); //close stream
      }
      else //Open failed
      {
          		std::cerr << "Failed to open /dev/urandom" << std::endl;
      }
      std::mt19937_64 mtrnd;
      std::seed_seq seed(&seeds[0], &seeds[624]);
      mtrnd.seed(seed);
      return mtrnd;
  }
  ```

Even with this modification, you may not be able to resolve the issue.
  In this case, please write a code to initialise mt19937\_64 in ``std::mt19937_64 init_random_number_generator()``.

### Uage:

[c, x, Q, q, p_vals] = km_config(A, num_of_runs, alpha, num_of_rand_nets);
 
 
  **INPUT:** 
 
  * `A` - N times N adjacency matrix, where N is the number of nodes. A(i, j) = w if nodes and j are adjacent, where w is the edge weight. Otherwise A(i, j) = 0. Matrix A should be symmetric, i.e., A(i, j) = A(j, i).
      
  * (optional) `num_of_runs` - Number of runs. (Default: 10) 
      
  * (optional) `alpha` - Statistical significance level before the Šidák correction. (Default: 1) If alpha is not set, the statistical test is not carried out. 
      
  * (optional) `num_of_rand_nets` - Number of randomised networks. (Default: 500)

**OUTPUT:**

* `c` - N-dimensional column vector. c(i) is the index of the core-periphery pair to which node i belongs.
          If node i belongs to an insignificant core-periphery pair, then c(i) = NaN.
      
  * `x` - N-dimensional column vector. If node i is a core node, x(i) = 1. If node i is a periphery node, x(i) = 0.
          If node i belongs to an insignificant core-periphery pair, then x(i) = NaN.
      
  * `Q` - The quality value of the detected core-periphery pair.
      
  * `q` - C-dimensional column vector. q(i) is the contribution of the i-th core-periphery pair to Q.
          C is the number of core-periphery pairs.
      
  * `p_vals` - C-dimensional column vector. p_vals(i) is the statistical significance of the i-th core-periphery pair.

### Example (src/matlab/example.m)
    
```matlab
% Construct a network with two core-periphery pairs 
A = [
	ones(5),ones(5);
	ones(5),zeros(5);
];
A = A - diag(diag(A));
A = kron(eye(2), A);

% Display adjacency matrix
disp('---- Adjacency matrix ----') 
full(A)

% Run KM-config algorithm
[c, x, Q, q] = km_config(A) % without statistical test

[c, x, Q, q, p_values] = km_config(A, 10, 0.05) % with statistical test
```

## Python 
      
### Compile:

To compile, type
        
  ```bash
  make python 
  ```
    
This creates a shared library ''src/python/km_config.cpython-36m-x86_64-linux-gnu.so'' callable from python. 
Copy the shared library to your working directory. 
 
### Uage:

```python
import km_config as kmconfig
cppairs = kmconfig.detect(edges, num_of_runs = 10, significance_level = 1, num_of_rand_nets = 500)
```
 
 
  **INPUT:** 
 
  * `edges` - Mx3 Numpy array, where M is the number of edges. Data type should be float (e.g., edges.astype(float)). The first and second columns indicate the IDs of nodes connected by an edge. The third column indicates the weight of the edge.
      
  * (optional) `num_of_runs` - Number of runs. (Default: 10) 
      
  * (optional) `significance_level` - Statistical significance level before the Šidák correction (Default: 1). If significance_level = 1, the statistical test is not carried out. 
      
  * (optional) `num_of_rand_nets` - Number of randomised networks. (Default: 500)

**OUTPUT:**

* `cppairs` - List of length 4. 
    * cppairs[0] - Numpy array of length N, where N is the number of nodes. cppairs[0][i] indicates the index of the core-periphery pair to which node i belongs.
    * cppairs[1] - Numpy array of length N. cppairs[1][i] = 1 or = 0 indicates a core or a peripheral node, respectively.
    * cppairs[2] - Numpy array of length N. cppairs[2][i] indicates the p-value of the core-periphery pair to which node i belongs.
    * cppairs[3] - Numpy array of length N. cppairs[3][i] = True or False indicates that node i belongs to the significant or insignificant cppairs, respectively.

### Example (src/python/example.py)
    
```python
import csv
import numpy as np
import km_config as kmconfig

linkfilename='example_edge_list.txt'
edges = np.genfromtxt(linkfilename, delimiter=' ', skip_header = 0)
cppairs = kmconfig.detect(edges.astype(float), significance_level = 0.05)

print(cppairs)
```

## Files
Directory src/lib contains the main code written in C++ 
  * src/lib/km_config.h is the header file.
  * src/lib/km_config.cpp is the implementation file of the header file, km_config.h.

Directory src/cpp contains C++ code for command-line cliend 
  * src/cpp/km_config_cl.cpp is the code for the command-line client.
  * src/cpp/example.m is a usage example.
  * src/cpp/example_edge_list.txt is an edge list of a network consisting of two idealised core-periphery pairs.
  * src/cpp/makefile is the makefile for the command-line client.

Directory src/matlab contains Matlab wrapper 
  * src/matlab/km_config_mex.cpp is the MATLAB wrapper for the C++ code (km_config.h and km_config.cpp)
  * src/matlab/km_config.m is the code for the MATLAB client.
  * src/matlab/example.m is a usage example.
  * src/matlab/makefile is the makefile for the MATLAB code.

Directory src/python contains Python wrapper 
  * src/python/km_config.cpp is the Python wrapper for the C++ code (km_config.h and km_config.cpp)
  * src/python/example.py is a usage example.
  * src/python/makefile is the makefile for the Python code. 
  * src/python/CMakeLists.txt is the cmake file for the Python code. 
  * src/python/example_edge_list.txt is an edge list of a network consisting of two idealised core-periphery pairs.

## Requirement
      
 * MATLAB 2012 or later.
 * Python3.4 or later
 * Cmake2.8 or later

---
Last updated: 17 October 2017

km_config

km_config Latest Version Download

Today's picks

Was this content helpful?

Recent Feedback

km_config

km_config Latest Version Download

Today's picks

Was this content helpful?

Recent Feedback

Download README