abd-clam 0.10.0-dev0

Clustered Learning of Approximate Manifolds
Documentation 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70

import argparse
import logging
import os
import pathlib

from abd_clam.utils import helpers

from . import download
from . import inference
from . import training

logging.basicConfig(
    format="%(asctime)s - %(levelname)s - %(name)s - %(funcName)s - %(message)s",
    datefmt="%d-%b-%y %H:%M:%S",
)
logger = helpers.make_logger(__name__)

DATA_ROOT = pathlib.Path(
    os.environ.get(
        "DATA_ROOT",
        pathlib.Path(__file__).parent.parent.parent.parent.joinpath("data"),
    ),
).resolve()
assert DATA_ROOT.exists(), f"DATA_ROOT not found: {DATA_ROOT}"

DATA_DIR = DATA_ROOT.joinpath("anomaly_data")
DATA_DIR.mkdir(exist_ok=True)

OUTPUT_DIR = DATA_ROOT.joinpath("trained_models")
OUTPUT_DIR.mkdir(exist_ok=True)

MODES = [
    "download",
    "training",
    "inference",
]

logger.info("Parsing arguments...")
parser = argparse.ArgumentParser(
    prog="chaoda",
    description="Run CHAODA in its various modes with default settings.",
)

""" Define the arguments """
parser.add_argument(
    "--mode",
    dest="mode",
    type=str,
    required=True,
    help=f"What mode to run. Must be one of {MODES}.",
)

args = parser.parse_args()

mode = args.mode
if mode not in MODES:
    msg = f"`mode` must be one of {MODES}. Got {mode} instead."
    raise ValueError(msg)

if mode == "download":
    download.download_and_save(DATA_DIR)
    download.load(DATA_DIR)
elif mode == "training":
    models_dir = OUTPUT_DIR.joinpath("pretrained_models")
    models_dir.mkdir(exist_ok=True)
    training.default_training(DATA_DIR, models_dir)
else:  # mode == 'inference'
    inference_dir = OUTPUT_DIR.joinpath("inference_results")
    inference_dir.mkdir(exist_ok=True)
    inference.run_inference(DATA_DIR, inference_dir)