Networkx is a very strang tool for generating graphs and it provides algorithms to analyze them http://networkx.github.io/

To me pajek is one of the best programs I have seen. I have also used igraph and Centibin

http://vlado.fmf.uni-lj.si/pub/networks/pajek/

http://igraph.sourceforge.net/

http://centibin.ipk-gatersleben.de/

I'm using R. It has plenty of packages with built-in functions and it is very easy to use. For example if you are analysing food webs, cheddar package and igraph are among the most common used ones.

My lab is using enaR, an R package for ecosystem network analysis. It builds on the network and sna packages. I like it because it lets me keep most of my analytical workflow within a single software.

R has many good packages. Cytoscape and all it's addins are excellent. Network 3D by Rich Williams is great for binary webs.

Thank you for the informative feedback. I am a frequent user of NetworkX, which is very comfortable and has a comprehensive collection of algorithms. However, NetworkX does not scale well to graphs with billions of edges. My own project, NetworKit, focuses on the analysis of such large graphs using parallel and approximation algorithms. Two projects that are likely to scale as well are graph-tool and igraph, which have a similar focus and architecture.

NodeXL might be a useful software; but i could not judge between Pajek and NodeXL, which one was the best. 

http://nodexl.codeplex.com/

Gephi and NetworkWorkbench

Gephi and NetLogo 

Hi

I think DIgSILENT PowerFactory is the best for your simulation

I prefer UCINET, Pajek, SIENA and ORA

I apply UCINET

I prefer Networkx. The main reason is that it is a python package and is suitable to work with other python packages such as matplotlib, scipy for more analysis, plots,....

I prefer Networkx. This is a python package and it's in continous developmente. I use this package frequently to analize small and big graphs. You can use it with other great packages such as matplotlib or scipy.

I use another tools like Gephi or Pajek but my main tool for SNA is NetworkX.

I apply Brain Connectivity Toolbox (BCT, http://www.brain-connectivity-toolbox.net/  source C++,  Matlab and Octave) contains a large selection of complex network measures, statistic and comparation, by Sporn/Rubinov.

I use another source: http://strategic.mit.edu/downloads.php?page=matlab_networks .

I prefer networkX because it is purely python implemented package and it is trivial to install.

I work complex networks like that:

UCINET: for medium size networks

PAJEK: for very large networks

ORA: for metanetworks

Jorge

For computational analysis and simulation - Igraph Package in R

For Visualization - Gephi

For Centrality measurements - PajekX

UCINET: for medium size networks

PAJEK: for very large networks

R packages Network, statnet, enaR and netindices

Python has powerful modules for complex network analysis.

I prefer UCINET for medium size network because is attractive and simple. However I think "R" is very powerful If you have practice with it.

I'd suggest Pyntacle, our python package for the identification of key-players and team-effects in networks. It exploits parallel and GPU computing for heavy computations on big graphs. Give us your feedbacks!

http://pyntacle.css-mendel.it

I prefer Pajek, Gephi, R, and NetworkX.

From the programming perspective, we can find the most popular software/library by statistical analysis of StackOverflow data available at https://data.stackexchange.com/

https://archive.org/download/stackexchange

I prefer R, Gephi, Cytoscape and Octave.

I'd prefer writing code with java

It's good for large datasets and you have power to do whatever you want

I am a heavy user of igraph library, especially its R version. The igraph library is very well designed and well maintained and the speed is quite fast on single core level. Although it lacks multi core support so far for some graph algorithm, but for me its not a big problem til now. Just as a reminder, there is a very good graph library "LightGraph.jl" for a quite new born programming language "Julia". It is born to support mutli-core functionality. The Julia has been recognized as a very promising scientific computing language with a very academia friendly design. If you need to perform network analysis on large network dataset, pls check the graph-tool library for Python. And it support openMP for parallel algorithm. For the mutlilayer network analysis, you can choose "multinet"(both c and R library are available) or two other R packages “multiplex” and "multigraph". I don't quite recommend networkX library, since its developed based on python and its quite slow compare to other libraries. Other packages in R: statnet(social network analysis, especially for ERGM model), tsna and networkDynamic for temporal network analysis. "netdiffuseR": Analysis of Diffusion and Contagion Processes on Networks.

I have started writing in java. I also use Gephi. Started learning R

I found sage (sagemath.org) to be useful for working with graphs.

I only use NetworkX in Python, which I think gives sufficient flexibility and computational power to carry out almost any task.

I'm using Python with some packages. It's a powerful tool.

I think that the best software for doing projects of complex networks are R and python. I almost most of my projects have done by those. fortunately, the library of the softwares is strong and updated.

Dear Christian L. Staudt ,

I agree with Kamal Berahmand in using R that has very useful packages like igraph. You can do a wide range of analysis and visualization by this package.

Hope it would help.

"https://graph-tool.skewed.de/performance"

Here are some comparison given between networkx and igraph library.

I am using networkX library in my project , and in this link also given extra information about its uses and benefits.

Use NetworkX for smaller networks and dynamic networks. NetworkX is pure Python, well documented and handles changes to the network gracefully.

iGraph is more performant in terms of speed and ram usage but less flexible for dynamic networks. iGraph is a C library with very smart indexing and storage approaches so you can load pretty large graphs in ram.

"https://www.reddit.com/r/Python/comments/4g9lp0/opinions_on_igraph_vs_netwrokx_in_python/"

Python programming Language with the Networkx library holds great prospect. that's what i use.. being a heavy python user

We use igraph for most functionalities. However, the plotting options in igraph leave a lot to be desired.

centrality metrics: igraph

community finding: igraph

plotting: ggnet or ggalley

ergm: sna/network

epidemiological modeling on dynamic networks: custom scripts written in R

contact tracing in dynamic networks: EpiContactTrace

A combination of Gephi, Matlab, SPSS, and MS Excel

Cytoscape software best for network analysis till now. It come in GUI and easy to handle.

Pajek, iGraph, Mathematica, and Gephi.

Is there a library in python that is for diffusion analysis in graphs? Something similar or better than those found in R?

To be able to deal with huge amount of data, you may consider networkit with Python. It is kind of a parallel and fast implementation of the networkx package. It is not as complete and as easy to use as Networkx, but it is very efficient !

I' m use the gephi for the visualization and igraph in R or Python for structural analysis. There is also the Networkx is very good!

According to Jure Leskovec paper in order to work with Network structural Analysis framework, he already analysed three of them based on running time and here's the result in order.

1.SNAP

2.iGraph

3.NetworkX

O igraph in R, acredito que seja mais fácil de utilizar. No entanto, isso fica critério do pesquisador

I often use igraph-R, igraph-python and NetworkX-python for network analysis (e.g., community detection, node evaluation).Gephi and matplotlib are good for the network visualization.

neo4j?

I almost use NetworkX package Developed for Python and Gephi

I used Igraph a bit.

I m using networkx and gephi you can use also i graph library 4 python. Also used anaconda ide for development and spyder as a tool.

I am using Gephi. But you have to install Java SE runtime enviroment prior to running the Gephi. Gephi is an Open-source software.

OPNET

Igraph (https://igraph.org) and Gephi (https://gephi.org)

R,

matlab

UCINET, PAJEK AND ORA

yEd software from german open-source project

An ecological analysis network can be applied to social, ecological, and complex socio-ecological systems. Very convent tool you have input-output tables (energy, environmentally extended economic data, etc.)

I am using R

NetworKit is a library to try out for complex graph analyses.

Pyntacle (pyntacle.css-mendel.it) if you are interested in group-centrality metrics

Opnet

Don't go for any tool that is slow. A lot of above mentioned tools are pretty slow. Here you have a nice comparison, which claims that Networkit (Python) is one of the fastest (it's not a surprise, it's written mostly in C, in the most vulnerable places). However, it's pretty slow with visualisation of huge graphs (which was not checked in the benchmark, but I'm saying that from my experience). In such cases I believe you can use Gephi, which has a graphics card support for visualisations.

https://www.timlrx.com/blog/benchmark-of-popular-graph-network-packages

Gephi for plot large networks.

NEA. m package for Matlab is good alternative to perform ecological network analysis. It allows to analyze even social and engineering network to determine where in supply chain of a product management interventions should be applied improve system efficiency and sustainability.

The major limitation, however, this package can be applied only to determined system at steady state.