def main():
args = parse_args()
backend = hp.LocalBackend(gsa_key_file=os.path.abspath("misc-270914-cb9992ec9b25.json")) if args.local else hp.BatchBackend(args.project)
p = hp.Pipeline(backend=backend, name=args.name)
for i, vcf in enumerate([
"gs://macarthurlab-rnaseq/grch38_vcfs/RGP_273_3_R1.SNPs.vcf.gz",
"gs://macarthurlab-rnaseq/grch38_vcfs/RGP_248_3.SNPs.vcf.gz",
"gs://macarthurlab-rnaseq/grch38_vcfs/RGP_54_3_2.SNPs.vcf.gz",
#"gs://macarthurlab-rnaseq/grch38_vcfs/RGP_7_1_2.SNPs.vcf.gz",
#"gs://macarthurlab-rnaseq/grch38_vcfs/RGP_7_2_2.SNPs.vcf.gz",
#"gs://macarthurlab-rnaseq/grch38_vcfs/RGP_7_3_2.SNPs.vcf.gz",
#"gs://macarthurlab-rnaseq/grch38_vcfs/RGP_7_4_2.SNPs.vcf.gz",
#"gs://macarthurlab-rnaseq/grch38_vcfs/RGP_7_5_2.SNPs.vcf.gz",
]):
t = p.new_task(name=f"olego{i}")
hg38_fasta = p.read_input_group(fa=os.path.expanduser("~/p1/ref/GRCh38/hg38.fa"), fai=os.path.expanduser("~/p1/ref/GRCh38/hg38.fa.fai"))
t.image("weisburd/olego:latest")
# switch to user account
t.command(f"gcloud auth activate-service-account --key-file /gsa-key/key.json")
t.command(f"gsutil -m cp -r gs://weisburd-misc/creds/.config /tmp/")
t.command(f"rm -rf ~/.config")
t.command(f"mv /tmp/.config ~/")
t.command(f"gcloud config set account [email protected]")
t.command(f"gcloud config set project seqr-project")
output_dir = os.path.basename(vcf).replace(".SNPs.vcf.gz", "")
#t.command(f"gsutil -m cp {hg38_fasta} .")
#t.command(f"gsutil -m cp {HG38_FASTA}.fai .")
t.command(f"gsutil -m cp {vcf}* .")
t.command(f"mkdir {output_dir}`")
t.command(f"create_genomes --fasta {hg38_fasta.fa} --vcf {os.path.basename(vcf)} --outdir {output_dir}")
t.command(f"gsutil -m cp -r {output_dir} gs://macarthurlab-rnaseq/grch38_personal_reference/")
#if args.local:
# p.write_output(t.ofile, 'temp.txt')
#else:
# p.write_output(t.ofile, 'gs://gnomad-bw2/temp.txt')
p.run()
break
#t.command(f"gsutil ls > {t.ofile}")
if isinstance(backend, hp.BatchBackend):
backend.close()
def stress():
p = pipeline.Pipeline(
name='stress',
backend=pipeline.BatchBackend(billing_project='hail'),
default_image='ubuntu:18.04')
for i in range(100):
t = (p.new_task(name=f'parent_{i}'))
d = random.choice(range(4))
if flip(0.2):
t.command(f'sleep {d}; exit 1')
else:
t.command(f'sleep {d}; echo parent {i}')
for j in range(10):
d = random.choice(range(4))
c = (p.new_task(name=f'child_{i}_{j}').command(
f'sleep {d}; echo child {i} {j}'))
c.depends_on(t)
if flip(0.2):
c._always_run = True
p.run(open=False, wait=False)
if __name__ == '__main__':
if len(sys.argv) != 6:
raise RuntimeError(f'usage: <script.py> DOCKER_IMAGE_URL BUCKET_BASE SHA N_REPLICATES N_ITERS')
BENCHMARK_IMAGE = sys.argv[1]
BUCKET_BASE = sys.argv[2]
SHA = sys.argv[3]
N_REPLICATES = int(sys.argv[4])
N_ITERS = int(sys.argv[5])
labeled_sha = SHA
label = os.environ.get('BENCHMARK_LABEL')
if label:
labeled_sha = f'{labeled_sha}-{label}'
p = pl.Pipeline(name=f'benchmark-{labeled_sha}',
backend=pl.BatchBackend(billing_project='hail'),
default_image=BENCHMARK_IMAGE,
default_storage='100G',
default_memory='7G',
default_cpu=2)
resource_tasks = {}
for r in all_resources:
t = p.new_task(f'create_resource_{r.name()}').cpu(4)
t.command(f'hail-bench create-resources --data-dir benchmark-resources --group {r.name()}')
t.command(f"time tar -cf {r.name()}.tar benchmark-resources/{r.name()} --exclude='*.crc'")
t.command(f'ls -lh {r.name()}.tar')
t.command(f'mv {r.name()}.tar {t.ofile}')
resource_tasks[r] = t
all_benchmarks = list_benchmarks()
assert len(all_benchmarks) > 0
from hailtop.hailctl.dev.benchmark.run.utils import list_benchmarks
if __name__ == '__main__':
if len(sys.argv) != 6:
raise RuntimeError(
f'usage: <script.py> DOCKER_IMAGE_URL BUCKET_BASE SHA N_REPLICATES N_ITERS'
)
BENCHMARK_IMAGE = sys.argv[1]
BUCKET_BASE = sys.argv[2]
SHA = sys.argv[3]
N_REPLICATES = int(sys.argv[4])
N_ITERS = int(sys.argv[5])
p = pl.Pipeline(name='benchmark',
backend=pl.BatchBackend(),
default_image=BENCHMARK_IMAGE,
default_storage='10G',
default_memory='7G',
default_cpu=2)
make_resources = p.new_task('create_resources').cpu(4)
make_resources.command(
'hailctl dev benchmark create-resources --data-dir benchmark-resources'
)
make_resources.command(
"time tar -czf benchmark-resources.tar.gz benchmark-resources --exclude='*.crc'"
)
make_resources.command('ls -lh benchmark-resources.tar.gz')
make_resources.command(
f'mv benchmark-resources.tar.gz {make_resources.ofile}')
all_benchmarks = list_benchmarks()
from hailtop.hailctl.dev.benchmark.run.utils import list_benchmarks
if __name__ == '__main__':
if len(sys.argv) != 6:
raise RuntimeError(
f'usage: <script.py> DOCKER_IMAGE_URL BUCKET_BASE SHA N_REPLICATES N_ITERS'
)
BENCHMARK_IMAGE = sys.argv[1]
BUCKET_BASE = sys.argv[2]
SHA = sys.argv[3]
N_REPLICATES = int(sys.argv[4])
N_ITERS = int(sys.argv[5])
p = pl.Pipeline(name='benchmark',
backend=pl.BatchBackend(url='https://batch.hail.is'),
default_image=BENCHMARK_IMAGE,
default_storage='10G',
default_memory='7G',
default_cpu=2)
make_resources = p.new_task('create_resources')
make_resources.command(
'hailctl dev benchmark create-resources --data-dir benchmark-resources'
)
make_resources.command(
"time tar -czvf benchmark-resources.tar.gz benchmark-resources --exclude='*.crc'"
)
make_resources.command('ls -lh benchmark-resources.tar.gz')
make_resources.command(
f'mv benchmark-resources.tar.gz {make_resources.ofile}')
all_benchmarks = list_benchmarks()