Metadata-Version: 2.4
Name: pathlims
Version: 2.0.dev0
Summary: A package to study the limits of pathlength and efficiency of complex networks.
Project-URL: Homepage, https://github.com/gorkazl/PathLims
Project-URL: Source Code, https://github.com/gorkazl/PathLims
Project-URL: Bug Tracker, https://github.com/gorkazl/PathLims/issues
Project-URL: Documentation, https://github.com/gorkazl/PahLims/blob/main/README.md
Author: Romain Brasselet
Author-email: Gorka Zamora-López <gorka@Zamora-Lopez.xyz>
License-Expression: Apache-2.0
License-File: LICENSE.txt
Keywords: complex networks,graph theory,network analysis
Classifier: Development Status :: 5 - Production/Stable
Classifier: Intended Audience :: Developers
Classifier: Intended Audience :: Education
Classifier: Intended Audience :: Science/Research
Classifier: License :: OSI Approved :: Apache Software License
Classifier: Operating System :: OS Independent
Classifier: Programming Language :: Python
Classifier: Programming Language :: Python :: 3 :: Only
Classifier: Topic :: Scientific/Engineering :: Bio-Informatics
Classifier: Topic :: Scientific/Engineering :: Information Analysis
Classifier: Topic :: Scientific/Engineering :: Mathematics
Classifier: Topic :: Scientific/Engineering :: Physics
Classifier: Topic :: Software Development :: Libraries :: Python Modules
Requires-Dist: numpy>=1.6
Requires-Dist: scipy
Description-Content-Type: text/markdown

# PathLims – Ultra-Short and Ultra-Long Network Generation

PathLims is a package to study and generate networks with largest and shortest possible average pathlength (or largest and smallest global efficiencies). Networks are treated as adjacency matrices, represented as 2D NumPy arrays.

[![pypi version](https://img.shields.io/pypi/v/pathlims?logo=pypi)](https://pypi.org/project/pathlims/)
[![PyPI Downloads](https://img.shields.io/pypi/dm/pathlims.svg?label=PyPI%20downloads)](
https://pypi.org/project/pathlims/)
[![Apache-2.0 License](https://img.shields.io/badge/license-Apache-blue.svg?style=flat)](http://choosealicense.com/licenses/Apache-2.0/)

 

The package contains two modules:

- *limits.py*: Analytic expressions to calulate the boundaries of average pathlength and global efficiency for networks of arbitrary size and number of links.
- *generators.py*: Functions to create ultra-short and ultra-long networks, of arbitrary size and number of links.

Visit the help of each module for a complete list of functions.

> **NOTE:** PathLims is fully compatible with [pyGAlib](https://github.com/gorkazl/pyGAlib), a library for graph analysis in Python/Numpy, but it can be used independently.

#### Reference and citation

For a complete description of the boundaries for the average pathlength and global efficiency, and for a illustration of ultra-short / ultra-long network generation, see:

- G. Zamora-López & R. Brasselet "[Sizing complex networks](https://doi.org/10.1038/s42005-019-0239-0)" *Comms. Phys.* **2**:144 (2019).

Please cite the above reference if you use PathLims. Results for some special cases (connected and undirected graphs) can also be found in: D. Barmpoutis & R.M. Murray "*Extremal Properties of Complex Networks*" arXiv:1104.5532v1 (2011); and L. Gulyas, et al. "*An Estimation of the Shortest and Largest Average Path Length in Graphs of Given Density*" arXiv:1101.2549v1 (2011).


### INSTALLATION

Installation of PathLims is simple, only the [pip](https://github.com/pypa/pip) package manager is needed. To check whether `pip` is installed, open a terminal and type:

	pip --help

> **NOTE**: If you use Anaconda (or any other third-party package manager), we recommend to install the dependencies (python>=3.6, numpy, scipy) into the target environment using Anaconda before installing PathLims. Otherwise, `pip` will download and install those packages directly from PyPI as well, and you won't be able to manage them from Acanconda.

#### Installing from PyPI 

PathLims is registered in the official *Python Package Index*, [PyPI](https://pypi.org/project/pathlims/) . To install, open a terminal window and type:

	python3 -m pip install pathlims

To confirm the installation, open an interactive session (e.g., IPython or a Notebook) and try to import the library by typing `import pathlims`.

#### Direct installation from GitHub 

If you have [git](https://git-scm.com) installed, you may like to install PathLims directly from its GitHub repository. Open a terminal and type:

	python3 -m pip install git+https://github.com/gorkazl/PathLims.git@master

This will only download and install the package (files in "*src/pathlims/*") into your current environment. If desired, additional files of the repository (e.g. the examples in the *Examples/* folder) should be downloaded manually. You can choose to install the version in another branch by replacing the '*@master*' at the end of the command by '*@branchname*' of the desired branch.

#### Installing PathLims in editable mode

If you want to install PathLims such that you can make changes to it "*on the fly*" then, visit its GitHub repository [https://github.com/gorkazl/PathLims/](https://github.com/gorkazl/PathLims/), select a branch and then click on the green "*<> Code*" button on the top right and select "Download ZIP" from the pop-up menu. Once downloaded, move the *zip* file to a target folder (e.g., "*~/Documents/myLibraries/*") and unzip the file. Open a terminal and `cd` to the resulting folder, e.g.,

	cd ~/Documents/myLibraries/PathLims-master/

Once on the path (make sure it contains the *pyproject.toml* file), type:

	python3 -m pip install -e .

Do not forget the "." at the end which means "*look for the pyproject.toml file in the current directory*." This will install PathLims such that every time changes are made to the package (located in the path chosen), these will be inmediately available. You may need to restart the IPython or Jupyter notebook session, though.


### HOW TO USE PathLims

Since PathLims depends on NumPy, it is recommended to import NumPy first. Although this is not necessary for loading the package, NumPy functionalities and array manipulation will be often needed. Try importing pathlims:

	>>> import numpy as np
	>>> import pathlims

> **NOTE**: Importing pathlims imports also the two main modules, *limits* and *generators* into the namespace. They can be called as `pathlims.limits`and `pathlims.generators`

##### Example 1 – Ultra-short graph

Let's generate an ultra-short graph of *N = 8* nodes and *L = 11* edges. We will use the random connectivity case:

	>>> import numpy as np
	>>> import pathlims
	>>> N = 8; L = 11
	>>> usnet = pathlims.generators.USgraph(N,L, uscase='Random')
	>>> print(usnet)
	array([[0, 1, 1, 1, 1, 1, 1, 1],
		[1, 0, 0, 1, 0, 0, 1, 0],
		[1, 0, 0, 0, 0, 0, 0, 0],
		[1, 1, 0, 0, 0, 0, 0, 0],
		[1, 0, 0, 0, 0, 0, 1, 0],
		[1, 0, 0, 0, 0, 0, 1, 0],
		[1, 1, 0, 0, 1, 1, 0, 0],
		[1, 0, 0, 0, 0, 0, 0, 0]], dtype=uint8)

The first node corresponds to the central hub in the initial star graph. The presence of this hub guarantees the diameter of the network to be diam(G) = 2. The remaining 4 edges are seeded at random.

![Figure1](Figs/USgraph_Rand.png)

Let's calculate, numerically, the average pathlength of this network. For this we use the `FloydWarshall()` function in the *helpers.py* module to calculate the pairwise distance matrix:

	>>> import pathlims.helpers
	>>> dij = pathlims.helpers.FloydWarshall(usnet)
	>>> avlen = ( dij.sum() - dij.trace() ) / ( N*(N-1) )
	>>> print( avlen )
	1.60714285714

The global efficiency is calculated as the average of the inverse of the distances:

	>>> eij = 1.0 / dij
	>>> effic = ( eij.sum() - eij.trace() ) / (N*(N-1))
	>>> print (effic )
	0.696428571429

We now corroborate that the numerical results match the theoretical estimations:

	>>> pathlims.limits.Pathlen_USgraph(N,L)
	1.6071428571428572
	>>> pathlims.limits.Effic_USgraph(N,L)
	0.6964285714285714

##### Example 2 – Sparse ultra-short digraphs (directed graph)

In the range *N ≤ L ≤ 2(N-1)* connected digraphs with shortest pathlength possible are characterised by a particular class of networks, **flower digraphs**, which consist of a small set of directed cycles all overlapping in a single node. We generate the flower digraph with *N = 8* nodes and *L = 11* arcs:

	>>> fdnet = pathlims.generators.USdigraph_FlowerDigraph(N,L)
	>>> fdnet
	array([[0, 1, 1, 0, 1, 0, 1, 0],
		[1, 0, 0, 0, 0, 0, 0, 0],
		[0, 0, 0, 1, 0, 0, 0, 0],
		[1, 0, 0, 0, 0, 0, 0, 0],
		[0, 0, 0, 0, 0, 1, 0, 0],
		[1, 0, 0, 0, 0, 0, 0, 0],
		[0, 0, 0, 0, 0, 0, 0, 1],
		[1, 0, 0, 0, 0, 0, 0, 0]], dtype=uint8)

![Figure1](Figs/FlowerDigraph.png)

The average pathlenth and gloal efficiency are numerically calculated as before:

	>>> dij = pathlims.helpers.FloydWarshall(fdnet)
	>>> avlen = ( dij.sum() - dij.trace() ) / ( N*(N-1) )
	>>> print(avlen)
	2.33928571429
	>>> eij = 1./dij
	>>> effic = ( eij.sum() - eij.trace() ) / (N*(N-1))
	>>> print(effic)
	0.5178571428571429

Finally, we corroborate that the numerical results match the theoretically expected values:

	>>> pathlims.limits.Pathlen_FlowerDigraph(N,L)
	2.3392857142857144
	>>> pathlims.limits.Effic_FlowerDigraph(N,L)
	0.51785714285714302

##### Data I/O

Since PathLims is based on NumPy arrays, saving and reading of adjacency matrices, can be performed using the usual data I/O functionalities of NumPy. See for example the documentation for functions: `loadtxt()`, `savetxt()`, `load()`, `save()` and `savez()`. See also the *tools.py* module in pyGAlib for other data conversions.



### HOW TO FIND FURTHER DOCUMENTATION

While working in an interactive session, after importing a module, the built-in `help()` function will show further details:

	>>> import modulename
	>>> help(modulename)

The help for PathLims (`help(pathlims)`) shows the general summary of the package and a list of all the modules in the library. The help for each module, e.g., `help(pathlims.limits)` or `help(pathlims.generators)` will display module specific information and a list of all the functions in the module.
For further details regarding each function, type:

	>>> help(pathlims.modulename.functionname)

IPython and Jupyter notebook users, the help command is replaced by a question mark after the module's or function's name, e.g.:

	>>> modulename?
	>>> functionname?

For questions, bug reports, etc, please write to <gorka@Zamora-Lopez.xyz>, or open an issue in GitHub.


### LICENSE
Copyright 2018, Gorka Zamora-López <<gorka@Zamora-Lopez.xyz>> and Romain Brasselet.

Licensed under the Apache License, Version 2.0. You may not use this software except in compliance with the License. You may obtain a copy of the License at

[http://www.apache.org/licenses/LICENSE-2.0](http://www.apache.org/licenses/LICENSE-2.0)

Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the specific language governing permissions and limitations under the License.


-----------------------------------------------------------------
### WHAT IS NEW

##### November 9, 2025 (Release of Version 2)

Stable version 2.0 checked, validated and released.

* The library has been reshaped to be compliant with the modern [PyPA specifications](https://packaging.python.org/en/latest/specifications/).
* [Hatch](https://hatch.pypa.io/latest/) was chosen as the tool to build and publish the package. See the *pyproject.toml* file. 
* Bug fixes to adapt to the various changes in NumPy since last release of PathLims.
* Sample and validation scripts in the "*Examples/*" folder revised and adapted to recent changes in Python and NumPy. 


