sparse-learning-CRN

Data-based sparsity learning of chemical reaction networks (CRNs).

These codes have been used to produce numerical results of the paper:

Learning chemical reaction networks from trajectory data, Siam J. Appl. Dyn. Syst., 18(4), pp. 2000-2046, 2019 .

Arxiv version is here.

DEPENDENCIES

This package relies on the following external libraries.

1. libconfig. This library is used to process the configuration file: ./working_dir/sparse_learning.cfg.

2. RANLIB.C (file: ranlib.c.tar.gz). This library provides random number generators. It is used in the code ./src/ssa.cpp to generate random trajectories of reactions.

To build a parallel code (optional), we also need

3. MPICH. The trajectory data will be distributed and analysed by different processors. Therefore, a parallel code is helpful, when there are multiple trajectories.

COMPILE & INSTALL

1. Install the above libraries, if necessary.

2. Download the source code.

	git clone https://github.com/zwpku/sparse-learning-CRN.git

The code should be avaiable in the directory ./sparse_learning_CRN

3. Enter the directory containing source files.

  	cd ./sparse_learning_CRN/src

4. Make sure paths of the directories containing the header and library files are provided.

5. Compile.

        make