Unverified 提交 8c6f9e15 authored 作者: Glenn Jocher's avatar Glenn Jocher 提交者: GitHub

Update `dataset_stats()` for zipped datasets (#3926)

* Update `dataset_stats()` for zipped datasets @KalenMike * cleanup
上级 850970e0
...@@ -888,9 +888,11 @@ def verify_image_label(args): ...@@ -888,9 +888,11 @@ def verify_image_label(args):
def dataset_stats(path='coco128.yaml', autodownload=False, verbose=False): def dataset_stats(path='coco128.yaml', autodownload=False, verbose=False):
""" Return dataset statistics dictionary with images and instances counts per split per class """ Return dataset statistics dictionary with images and instances counts per split per class
Usage: from utils.datasets import *; dataset_stats('coco128.yaml', verbose=True) Usage1: from utils.datasets import *; dataset_stats('coco128.yaml', verbose=True)
Usage2: from utils.datasets import *; dataset_stats('../datasets/coco128.zip', verbose=True)
Arguments Arguments
path: Path to data.yaml path: Path to data.yaml or data.zip (with data.yaml inside data.zip)
autodownload: Attempt to download dataset if not found locally autodownload: Attempt to download dataset if not found locally
verbose: Print stats dictionary verbose: Print stats dictionary
""" """
...@@ -899,8 +901,20 @@ def dataset_stats(path='coco128.yaml', autodownload=False, verbose=False): ...@@ -899,8 +901,20 @@ def dataset_stats(path='coco128.yaml', autodownload=False, verbose=False):
# Update labels to integer class and 6 decimal place floats # Update labels to integer class and 6 decimal place floats
return [[int(c), *[round(x, 6) for x in points]] for c, *points in labels] return [[int(c), *[round(x, 6) for x in points]] for c, *points in labels]
with open(check_file(path)) as f: def unzip(path):
# Unzip data.zip TODO: CONSTRAINT: path/to/abc.zip MUST unzip to 'path/to/abc/'
if str(path).endswith('.zip'): # path is data.zip
assert os.system(f'unzip -q {path} -d {path.parent}') == 0, f'Error unzipping {path}'
data_dir = path.with_suffix('') # dataset directory
return True, data_dir, list(data_dir.rglob('*.yaml'))[0] # zipped, data_dir, yaml_path
else: # path is data.yaml
return False, None, path
zipped, data_dir, yaml_path = unzip(Path(path))
with open(check_file(yaml_path)) as f:
data = yaml.safe_load(f) # data dict data = yaml.safe_load(f) # data dict
if zipped:
data['path'] = data_dir # TODO: should this be dir.resolve()?
check_dataset(data, autodownload) # download dataset if missing check_dataset(data, autodownload) # download dataset if missing
nc = data['nc'] # number of classes nc = data['nc'] # number of classes
stats = {'nc': nc, 'names': data['names']} # statistics dictionary stats = {'nc': nc, 'names': data['names']} # statistics dictionary
......
Markdown 格式
0%
您添加了 0 到此讨论。请谨慎行事。
请先完成此评论的编辑!
注册 或者 后发表评论