This version of the greedy set cover is a well optimized linear version.
The input format is
10 5
5 7 8 9 10
5 7
1 3 6 10
4 6 7 9 10
1 2 8 10
10 = elements 5 = sets Every line represents one set by the ids
git clone https://github.com/martin-steinegger/setcover
cd setcover
make
./setcover testdatafile
The program can be called like ./setcover filename
Example output
$ ./setcover testdatafile
start readin --- Done
init setcover --- Done
execute setcover --- Done
validate result VALID --- Done
write result setcover
set id 4 Element: 4, 6, 7, 9, 10,
set id 5 Element: 1, 2, 8,
set id 1 Element: 5,
set id 3 Element: 3,
The calculation for 40.000.000 sets with avg. 10 elements per set takes around 4 Minutes on a m2.2xlarge instance.
