High dimensional peaks finding via topological data analysis.
git clone https://github.com/fightinggggg/HDPF.git
cd HDPF
g++ PersistentHomology.cpp -std=c++11 -o ph
The input file should be a plain text file in tsv format with n columns. With first n-1 columns denotes the coordinates while the nth column denotes the cooresponding probability landscape. One thing to remember that the coordiantes must be integers while the probability can be float numbers.
cat test.txt
./ph test.txt
The output pf HDPF is a plain text in tsv format with 3 columns. Each column denotes the birth, death and lenght of persistent barcode.
The HDPF program relies on Eigen library to perform matrix calculation. If you find HDPF helpful in your research, please do not forget to cite Eigen at the same time.
Eigen: https://eigen.tuxfamily.org/