This software provides both a binary and a library (see note) for an anytime behavior model of multi-objective combinatorial optimization algorithms that, at each iteration, collect an efficient solution that maximizes the hypervolume contribution. It is assumed that all objective functions are to be maximized.
Note: the current software is intended to be used as a binary for now. The library API is not yet properly defined and no documentation is provided for now.
Install the latest binary using cargo with:
sh
cargo install moco-abm-bin
or compile from source with:
sh
cargo build --release
```
USAGE:
moco-abm [OPTIONS] -n
FLAGS: -h, --help Prints help information -V, --version Prints version information
OPTIONS:
-f
The input file should contain at least one segment in the following format
u1 u2 v1 v2
where $(u1, u2)$ and $(v1, v2)$ denote the coordinates of the linear segment endpoints. Multiple segments, and coordinates within the segments, can be separated by any whitespace.
Note: Points in the segments must be provided such that v1 > u1 and v2 < u2. Moreover, when multiple segments are provided, e.g.:
u1 u2 v1 v2
p1 p2 q1 q2
it is required that p1 >= v1 and that p2 <= v2.
Example of a valid segments list file:
0.0 1.0 0.7 0.7
0.7 0.7 1.0 0.0
The output is returned to stdout and consists of a .tsv with the following fields
| field | description | |:----------------|:---------------------------------------------| | index | index of the current point (starts at 1) | | hvcontribution | hypervolume contribution of this point | | hvcurrent | hypervolume of all returned points up to now | | hvrelative | currenthv relative to maximal hypervolume | | point | comma separated coordinates of the point |
Add this to your Cargo.toml:
[dependencies]
moco_abm = "0.1"
and this to your crate root:
rust
extern crate moco_abm;