required=True,
help="list of .npy filenames for labels to be split randomly e.g. "
"'--label_filenames training_y_0.csv validation_y_0.csv "
"training_y_-2.csv validation_y_-2.csv'. "
"Assumed to be grouped by threshold, in the same order as "
"the thresholds parameter.")
required_named.add_argument("--thresholds",
type=int,
nargs='+',
required=True,
help="list of thresholds to use to reduce dataset by similarity "
"score e.g. --thresholds 0 -2")
args = parser.parse_args()
log_utils.setup_logging(args.verbosity)
label_filenames = args.label_filenames
data_filenames = args.data_filenames
group_proportions = [80, 20]
thresholds = args.thresholds
logging.info(f"Reading in data frame {args.input}")
data_frame = utils.read_bound_pairs(args.input)
logging.info(f"Read {len(data_frame)} rows")
positives_df = data_frame[data_frame['binding_observed'] == 1].iloc[:args.num_negatives]
logging.info(f"Chosen {len(positives_df)} random positive samples")
label_filenames_grouped = [label_filenames[2*i:2*i+2] for i in range(len(thresholds))]
logging.info(f"{len(label_filenames)} label files grouped into "
"""Wrapper for snakemake to call the find_unique_bound_pairs script"""
import logging
import traceback
import peptidebinding.find_unique_bound_pairs as find_unique_bound_pairs
import peptidebinding.helper.log_utils as log_utils
log_utils.setup_logging(3, logfile=snakemake.log[0])
try:
find_unique_bound_pairs.main(
bound_pairs_tables=snakemake.input.bound_pairs,
output_file=snakemake.output.bound_pairs,
fragment_lengths_out=snakemake.output.fragment_lengths)
except:
logging.error(f"Unexpected error:\n{traceback.format_exc()}")
raise