flatspin-run-sweep

Run flatspin sweep.

usage: flatspin-run-sweep [-h] [-r {local,dist,none}] [-p PARAM] [-s key=SPEC]
                          [-n N] [-ns key=SPEC] [--seed SEED] [-m MODEL]
                          [-e ENCODER] [-o BASEPATH] [-f {csv,hdf,npy,npz}]
                          [--params-file PARAMS_FILE]
                          [--sweep-file SWEEP_FILE] [--list-params]
                          [--list-encoders] [--help-encoder] [-V]
                          [--show-plot]

-h, --help

show this help message and exit

-r {local,dist,none}, --run {local,dist,none}

run locally or distributed on a cluster

-p, --param

set model parameter key=value

-s, --sweep

set sweep parameter key=SPEC

-n <n>, --repeat <n>

repeat each experiment N times (default: 1)

-ns, --repeat-spec

repeat each according to key=SPEC

--seed <seed>

random seed

-m <model>, --model <model>

name of model (default: SquareSpinIceClosed)

-e <encoder>, --encoder <encoder>

input encoder (type of external field, default: Sine)

-o <basepath>, --basepath <basepath>

output directory for results

-f {csv,hdf,npy,npz}, --format {csv,hdf,npy,npz}

format of output files (default: npz)

--params-file <params_file>

load parameters from file (params.csv), loaded before -p

--sweep-file <sweep_file>

load sweep (index.csv) from file

--list-params

--list-encoders

--help-encoder

-V, --version

display flatspin version

--show-plot

plot the state as the simulation runs