def individuals_location(j):
j.generator(
RepeatG(
Product(
alpha_v=[0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9, 0.99]),
no_of_runs,
))
j.processor(Threading(config.User.get("cpu_count", os.cpu_count())))
j.runner(
SubprocessRunner.factory(
j.dependencies[build_individuals_location].artefacts["executable"],
input=ProcessArgs(P("alpha_v")),
output=Stdout(),
))
coll = j.collector(
Demux(["alpha_v"],
Factory(
Concatenate,
file_path=Join("individuals_location.out.", P("alpha_v"),
".tsv"),
)))
j.artefact(result=coll.aggregate)
j.after(
lambda j: print("Wrote results to",
list(map(lambda f: f.name, j.artefacts["result"]))))
j.after(lambda j: list(map(lambda f: f.close(), j.artefacts["result"])))
def cluster_impurity(j):
j.generator(
RepeatG(
Product(
# k=[30, 100, 500, 1000],
k=[500],
# k_over_l=[3, 5, 10, 20],
k_over_l=[20],
hhc=[0.1, 0.3, 0.5, 0.7, 0.9],
),
no_of_runs,
))
j.processor(Threading(config.User.get("cpu_count", os.cpu_count())))
j.runner(
SubprocessRunner.factory(
j.dependencies[build_cluster_impurity].artefacts["executable"],
input=ProcessArgs(P("k"),
Function(lambda p: p["k"] // p["k_over_l"]),
P("hhc")),
output=Stdout(),
))
coll = j.collector(
Demux(["k", "k_over_l", "hhc"],
Factory(
CSV,
file_path=Join("cluster_impurity.out.", P("k"), ".",
P("k_over_l"), ".", P("hhc"), ".tsv"),
input_csv_args=dict(delimiter=' '),
output_csv_args=dict(delimiter=' '),
)))
j.artefact(result=coll.aggregate)
j.after(
lambda j: print("Wrote results to",
list(map(lambda f: f.name, j.artefacts["result"]))))
j.after(lambda j: list(map(lambda f: f.close(), j.artefacts["result"])))
def subcluster_individuals(j):
j.generator(
RepeatG(
Product(
# k=[30, 100, 500, 1000],
k=[500],
# n=[500, 1000, 2000, 5000],
n=[5000],
alpha_v=[0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9, 0.99]),
no_of_runs,
))
j.processor(Threading(config.User.get("cpu_count", os.cpu_count())))
j.runner(
SubprocessRunner.factory(
j.dependencies[build_subcluster_individuals].
artefacts["executable"],
input=ProcessArgs(P("k"), P("n"), P("alpha_v")),
output=Stdout(),
))
coll = j.collector(
Demux(["k", "n", "alpha_v"],
Factory(
CSV,
file_path=Join("subcluster_individuals.out.", P("k"), ".",
P("n"), ".", P("alpha_v"), ".tsv"),
input_csv_args=dict(delimiter=' '),
output_csv_args=dict(delimiter=' '),
)))
j.artefact(result=coll.aggregate)
j.after(
lambda j: print("Wrote results to",
list(map(lambda f: f.name, j.artefacts["result"]))))
j.after(lambda j: list(map(lambda f: f.close(), j.artefacts["result"])))