# libigl - A simple C++ geometry processing library

https://github.com/libigl/libigl/

Get started with:

git clone --recursive https://github.com/libigl/libigl.git


libigl is a simple C++ geometry processing library. We have a wide functionality including construction of sparse discrete differential geometry operators and finite-elements matrices such as the cotangent Laplacian and diagonalized mass matrix, simple facet and edge-based topology data structures, mesh-viewing utilities for OpenGL and GLSL, and many core functions for matrix manipulation which make Eigen feel a lot more like MATLAB.

It is a header-only library. You do not need to compile anything to use, just include igl headers (e.g. #include <igl/cotmatrix.h>) and run. Each header file contains a single function (e.g. igl/cotmatrix.h contains igl::cotmatrix()). Most are tailored to operate on a generic triangle mesh stored in an n-by–3 matrix of vertex positions V and an m-by–3 matrix of triangle indices F.

Optionally the library may also be pre-compiled into a statically linked library, for faster compile times with your projects. This only effects compile time (run-time performance and behavior is identical). If in doubt, use the header-only default mode: (i.e. just include the headers you want to use).

We use the Eigen library heavily in our code. Our group prototypes a lot in MATLAB, and we have a useful MATLAB to libigl+Eigen conversion table.

We regularly test compiling our library on Mac OS X with clang, Linux with gcc and Windows with Visual Studio 2015 Community Edition.

## Tutorial

As of version 1.0, libigl includes an introductory tutorial that covers many functionalities.

## Installation

Libigl is a header-only library. You do not need to build anything to install. Simply add libigl/include to your include path and include relevant headers. Here is a small “Hello, World” program:

#include <igl/cotmatrix.h>
#include <Eigen/Dense>
#include <Eigen/Sparse>
#include <iostream>
int main()
{
Eigen::MatrixXd V(4,2);
V<<0,0,
1,0,
1,1,
0,1;
Eigen::MatrixXi F(2,3);
F<<0,1,2,
0,2,3;
Eigen::SparseMatrix<double> L;
igl::cotmatrix(V,F,L);
std::cout<<"Hello, mesh: "<<std::endl<<L*V<<std::endl;
return 0;
}


If you save this in hello.cpp, then you could compile this with (assuming Eigen is installed in /usr/local/include/eigen3):

g++ -std=c++11 -I/usr/local/include/eigen3 -I./libigl/include/ hello.cpp -o hello


Running ./hello would then produce

Hello, mesh:
0.5  0.5
-0.5  0.5
-0.5 -0.5
0.5 -0.5


## Dependencies

Dependencies are on a per-include basis and the majority of the functions in libigl depends only on the Eigen library.

### Optional dependencies

Libigl compartmentalizes its optional dependences via its directory organization in the include/ folder. All header files located directly in the include/igl/ folder have only stl and Eigen as dependencies. For example, all of the headers that depend on CGAL are located in include/igl/cgal. For a full list of optional dependencies check optional/CMakeLists.txt.

### GCC and the optional CGAL dependency

The include/igl/cgal/*.h headers depend on CGAL. It has come to our attention that CGAL does not work properly with GCC 4.8. To the best of our knowledge, GCC 4.7 and clang will work correctly.

### OpenMP and Windows

Some of our functions will take advantage of OpenMP if available. However, it has come to our attention that Visual Studio + Eigen + OpenMP does not work properly. Since we use OpenMP only to improve performance, we recommend avoiding OpenMP on Windows or proceeding with caution.

You can keep up to date by cloning a read-only copy of our GitHub repository.

## Known Issues

We rely heavily on Eigen. Nearly all inputs and outputs are Eigen matrices of some kind. However, we currently only officially support Eigen’s default column-major ordering. That means, we do not expect our code to work for matrices using the Eigen::RowMajor flag. If you can, change definitions like:

Eigen::Matrix<double, Eigen::Dynamic, 3, Eigen::RowMajor> A;


to

Eigen::Matrix<double, Eigen::Dynamic, 3, Eigen::ColMajor> A;
// or simply
Eigen::Matrix<double, Eigen::Dynamic, 3> A;


We hope to fix this, or at least identify which functions are safe (many of them probably work just fine). This requires setting up unit testing, which is a major todo for our development.

## Git Submodules

Libigl uses git submodules for its optional dependencies, in particular, those needed by the OpenGL viewer to run the examples in the tutorial. Git submodules allow use to treat clones of other libraries as sub-directories within ours while separating our commits. Read the documentation for a detailed explanation, but essentially our libigl repo stores a hash for each of its subrepos containing which version to update to. When a change is introduced in a dependencies repo we can incorporate that change by pulling in our sub-repo and updating (i.e. committing) that change to the hash.

When pulling new changes to libigl it’s also a good idea to update changes to subrepos:

git pull
git submodule update --recursive


## Unit testing

Libigl maintains separate repository for unit testing.

## How to contribute

libigl is primarily MPL2 licensed (FAQ). Some files contain third-party code under other licenses. We’re currently in the processes of identifying these and marking appropriately.

If you use libigl in your academic projects, please cite the papers we implement as appropriate. To cite the library in general, you could use this BibTeX entry:

@misc{libigl,
title = {{libigl}: A simple {C++} geometry processing library},
author = {Alec Jacobson and Daniele Panozzo and others},
note = {http://libigl.github.io/libigl/},
year = {2016},
}


## Projects/Universities using libigl

Libigl is used by many research groups around the world. In 2015, it won the Eurographics/ACM Symposium on Geometry Processing software award. Here are a few labs/companies/institutions using libigl: