Command Line Mode

Command Line Mode #

Except using EMOC in the GUI mode, we also provide command line (CMD) mode for these users who don’t have a display screen for some reasons(like on a remote server).

A Simple Case #

Go to the directory of executable file and open the terminal. Users can run EMOC with default parameters by just:

./EMOC

For windows, change the ’./EMOC’ to ‘EMOC.exe’. Some settings and results information will be printed in the terminal:

image

(Note: If you want to calculate the indicator values of optimization results, the executable file should in the same directory with ’/pf_data’.)

After executing, the optimized population results will be saved in /output/test_module/ directory.

Run EMOC with Different Parameters #

In CMD mode, users can also set different parameters of this run, an example is shown below:

./EMOC --algorithm MOEADDE --problem DTLZ1 --evaluation 25000 -N 100

this command sets the algorithm to MOEADDE, the problem to DTLZ1, the max evaluation to 25000 and the population number to 100. All acceptable parameters are listed in the following table:

Parameter NameDescriptionDefault Value
-h, --helpPrint some helpful information in the terminalNone
-g, --guiRun EMOC in GUI modeFalse
-a, --algorithmAlgorithm nameNSGA2
-p, --problemProblem nameZDT1
-N, --popPopulation size100
-m, --objProblem’s objective dimension2
-n, --decProblem’s decision variable dimension30
-e, --evaluationMax evaluation number for each run25000
-i, --intervalPopulation save interval in generation100
-r, --runThe number of runs1
--multithreadWhether to use multi-threadFalse
-t, --threadThread number (only valid when using multi-thread)4

Available Algorithms and Problems #

The available algorithms and problems in EMOC are listed below.

Algorithms:

Decomposition BasedDominance BasedIndicator BasedConstraintSingle Objective
MOEADNSGA2IBEACNSGA2GA
MOEADDENSGA3HypECMOEADDifferentialEvolution
MOEADDRASPEA2SMSEMOACTAEASA
MOEADGRASPEA2SDE
MOEADIRAtDEA
ENSMOEAD
MOEADCDE
MOEADSTM
MOEADPAS
MOEADM2M
MOEADD
MOEADDYTS
MOEADFRRMAB
MOEADUCB
RVEA

Problems:

Single ObjectiveMulti ObjectiveMany ObjectiveConstraint
SphereZDT SeriesDTLZ SeriesCDTLZ Series
AckleyUF SeriesMinusDTLZ SeriesDCDTLZ Series
RastriginBT SeriesMDTLZ Series
TSPMOEADDE_F SeriesWFG Series
KnapsackIMMOEA_F SeriesLSMOP Series
MOEADM2M_F Series