-
Notifications
You must be signed in to change notification settings - Fork 6
/
README.Rmd
105 lines (79 loc) · 4.62 KB
/
README.Rmd
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
---
output: github_document
---
# netgen: Generating random (clustered) networks in R
<!-- badges: start -->
[](http://cran.r-project.org/web/packages/netgen)
[](http://cran.rstudio.com/web/packages/netgen/index.html)
[](https://github.com/jakobbossek/netgen/actions)
[](https://codecov.io/gh/jakobbossek/netgen?branch=master)
<!-- badges: end -->
## Description
The *netgen* package offers several methods to generate random graphs respectively
networks for benchmarking purposes. You may want to generate some random graphs
or clustered graphs to perform benchmark studies on some fine algorithms for
the travelling salesperson problem as an example. Examples of generated networks
are depicted in the following figure.
## Installation
The package has recently been submitted to CRAN. Find additional information on the [official CRAN website](http://cran.r-project.org/web/packages/netgen/).
To install the current developement version of the package, install the [devtools package](http://cran.r-project.org/web/packages/devtools/index.html) by Hadley Wickham, load it and type the following line to a R session:
```splus
install_github("jakobbossek/netgen")
```
## Quickstart
To generate a random network in the euclidean plane [0,1] x [0,1] we simply
load the package and call the `generateRandomNetwork` function, providing the
desired number of nodes.
```splus
library(netgen)
set.seed(123)
x = generateRandomNetwork(n.points = 50L)
print(x)
autoplot(x)
```
To generate a network with points distributed around several clusters in the
[0,100] x [0, 100] space make use of the more powerful `generateClusteredNetwork`
function. It is mandatory to provide the number of points `n.points` and the number of clusters `n.cluster`.
**Note:** See the documentation of the function to read about how it works internally.
```splus
library(netgen)
set.seed(1)
x = generateClusteredNetwork(n.cluster = 5L, n.points = 150L, upper = 100)
print(x)
autoplot(x)
```
## News
### netgen v1.3 (Release data: 2016-01-16):
* Default number of digits for exportToTSPlibFormat is 10 now
* Added function isEuclidean
* rescaleNetwork now supports instances with depots
* rescaleNetwork now recomputes distance matrix
* rescaleNetwork throws error on non-euclidean networks
* Added function getTSPInstancesOverview to get an overview of files in TSPlib
format
* Added function filterTSPInstances which allows to subset/filter TSP instances
in TSPlib format by properties.
* Networks have an additional property edge.weight.type now (see TSPlib documentation
for details)
* Added function getValidEdgeWeightTypes
### netgen v1.2 (Release data: 2015-11-24):
* last line of TSPlib export now containts EOF string
* function importFromTSPlibFormat now has an additional argument round.distances with default value
TRUE which decides whether distances of EUC_2D instances should be rounded to the nearest integer value
### netgen v1.1 (Release data: 2015-05-19):
* added rescaleNetwork
* network objects now store the distance matrix internally
* network objects now have name and comment field
* lower and upper bound are now optional for makeNetwork/makeClusteredNetwork
* added possibility to pass point matching or point matching algorithm to morphInstances
* getOptimalPointAssignment now has another parameter method (new method to
solve the assigment to optimality)
* importFromTSPlibFormat and exportToTSPlibFormat for storing networks in the TSPlib format
* importFromFile and exportToFile for storing networks in proprietary format (similar to TSPlib format, but with additional information like cluster membership or arrival/request times)
* generate{Random,Clustered,Grid}Network now automatically generate a random name if none is provided
* there is no makeClusteredNetwork function anymore. Instead makeNetwork has an optional paramter membership
### netgen v1.0 (Release date: 2015-02-03):
* First submission to CRAN.
## Contact
Please address questions and missing features about the **netgen package** to the author Jakob Bossek <[email protected]>. Found some nasty bugs? Please use the [issue tracker](https://github.com/jbossek/netgen/issues) for this. Pay attention to explain the problem as good as possible. At its best you provide an example, so I can reproduce your problem.