This is the experimental code of paper
"Improved Approximation Algorithms for Capacitated Location Routing".

Environment:

1. language: c++;
2. cmake 3.15+;
3. g++ 11.4.0+.

Code organization:

1. instance: Benchmark. Note that the cost in the instances P111112-P133222 is multiplied by 10;
2. CMakeList.txt: Cmake file;
3. constant.h: Some static variables;
4. factory: Greedily open depots;
5. graph: Read data from file and build weight matrix;
6. main.cpp: Main file;
7. ps: Path-Alg. See the paper for details;
8. tour: Match and shortcut;
9. ts: Tree-Alg. See the paper for details;
10. tsp: Finding a tsp using 3/2 approximation algorithm.

Build and Execution:

1. You can use g++ directly or cmake tools to compile and run,
   before that please make sure your environment is linux/unix
   and satisfy the minimum requirements.
2. Using "mkdir build && cd build && cmake .. && make"
   to generate executable file, i.e., clr.
3. clr has two parameters: the first is dataFileName, please put dataFile
   in "instance" folder which is same directory as clr; the second
   parameter is Alpha. See the paper for details of parameter Alpha. 
   For using Alpha as an integer, we use Alpha*100 in our program.
4. Execution: "./clr dataFileName Alpha*100", e.g., "./clr P111112 100",
   where P111112 locates in "./instance/P111112" and it means Alpha=1.

OutPut:

1. The output of clr are "dataFileName Tree-AlgResult Path-AlgResult Time(ns)".
   You can change it in "main.cpp" or other files.