pytorch/examples
Ver no GitHubDoc comment on `accuracy` method in imagenet example, incorrect?
Open
#312 aberto em 27 de fev. de 2018
good first issue
Métricas do repositório
- Stars
- (21.634 stars)
- Métricas de merge de PR
- (Nenhuma PRs mesclada em 30d)
Description
I'm confused with the doc comment for the accuracy function in the imagenet example:
def accuracy(output, target, topk=(1,)):
"""Computes the precision@k for the specified values of k"""
maxk = max(topk)
batch_size = target.size(0)
_, pred = output.topk(maxk, 1, True, True)
pred = pred.t()
correct = pred.eq(target.view(1, -1).expand_as(pred))
res = []
for k in topk:
correct_k = correct[:k].view(-1).float().sum(0, keepdim=True)
res.append(correct_k.mul_(100.0 / batch_size))
return res
https://github.com/pytorch/examples/blob/master/imagenet/main.py#L298-L299
This seems like it computes accuracy and not precision as false positives are not accounted for. Should the doc comment read "Computes the accuracy@k for the specified values of k" or is my understanding of precision for object detection incorrect?
Many thanks for pytorch, it's a great library!