File size: 4,234 Bytes
c96678c |
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 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 |
import os, sys
import unittest
from pathlib import Path
import pandas as pd
sys.path.insert(0, os.path.abspath("."))
########### MODIFY SESSION SETTINGS BEFORE GLOBAL CONFIG IS IMPORTED #########
from acodet.create_session_file import create_session_file
create_session_file()
import json
with open("acodet/src/tmp_session.json", "r") as f:
session = json.load(f)
session["sound_files_source"] = "tests/test_files/test_audio_files"
session[
"generated_annotation_source"
] = "tests/test_files/test_generated_annotations"
session[
"annotation_destination"
] = "tests/test_files/test_combined_annotations"
session[
"generated_annotations_folder"
] = "tests/test_files/test_generated_annotations"
session[
"reviewed_annotation_source"
] = "tests/test_files/test_generated_annotations"
session["tfrecords_destination_folder"] = "tests/test_files/test_tfrecords"
with open("acodet/src/tmp_session.json", "w") as f:
json.dump(session, f)
##############################################################################
from acodet.annotate import run_annotation, filter_annots_by_thresh
from acodet.funcs import return_windowed_file, get_train_set_size
from acodet.models import GoogleMod
from acodet.combine_annotations import generate_final_annotations
from acodet.tfrec import write_tfrec_dataset
from acodet.train import run_training
from acodet import global_config as conf
class TestDetection(unittest.TestCase):
def test_annotation(self):
self.time_stamp = run_annotation()
df = pd.read_csv(
(
Path(conf.GEN_ANNOTS_DIR)
.joinpath(self.time_stamp)
.joinpath("stats.csv")
)
)
self.assertEqual(
df["number of predictions with thresh>0.8"][0],
326,
"Number of predictions is not what it should be.",
)
filter_annots_by_thresh(self.time_stamp)
file = list(
Path(conf.GEN_ANNOT_SRC)
.joinpath(self.time_stamp)
.joinpath(f"thresh_{conf.THRESH}")
.glob("**/*.txt")
)[0]
df = pd.read_csv(file)
self.assertEqual(
len(df),
309,
"Number of predictions from filtered thresholds " "is incorrect.",
)
class TestTraining(unittest.TestCase):
def test_model_load(self):
model = GoogleMod(load_g_ckpt=False).model
self.assertGreater(len(model.layers), 15)
# def test_tfrecord_loading(self):
# data_dir = list(Path(conf.TFREC_DESTINATION).iterdir())
# n_train, n_noise = get_train_set_size(data_dir)
# self.assertEqual(n_train, 517)
# self.assertEqual(n_noise, 42)
class TestTFRecordCreation(unittest.TestCase):
def test_tfrecord(self):
time_stamp = list(Path(conf.ANNOT_DEST).iterdir())[-1]
write_tfrec_dataset(annot_dir=time_stamp, active_learning=False)
metadata_file_path = Path(conf.TFREC_DESTINATION).joinpath(
"dataset_meta_train.json"
)
self.assertEqual(
metadata_file_path.exists(),
1,
"TFRecords metadata file was not created.",
)
with open(metadata_file_path, "r") as f:
data = json.load(f)
self.assertEqual(
data["dataset"]["size"]["train"],
517,
"TFRecords files has wrong number of datapoints.",
)
def test_combined_annotation(self):
generate_final_annotations(active_learning=False)
time_stamp = list(Path(conf.GEN_ANNOTS_DIR).iterdir())[-1].stem
combined_annots_path = (
Path(conf.ANNOT_DEST)
.joinpath(time_stamp)
.joinpath("combined_annotations.csv")
)
self.assertEqual(
combined_annots_path.exists(),
1,
"csv file containing combined_annotations does not exist.",
)
df = pd.read_csv(combined_annots_path)
self.assertEqual(
df.start.iloc[-1],
1795.2825,
"The annotations in combined_annotations.csv don't seem to be identical",
)
if __name__ == "__main__":
unittest.main()
|