all_metrics.py

3 F541: f-string is missing placeholders
2 F401: 'gds.common.utils.avg_over_groups' imported but unused (and 1 similar)
- 'gds.common.utils.avg_over_groups' imported but unused:
  8 from gds.common.utils import avg_over_groups, minimum, maximum, get_counts
- 'gds.common.utils.maximum' imported but unused:
  8 from gds.common.utils import avg_over_groups, minimum, maximum, get_counts
1 F841: local variable 'acc' is assigned to but never used
- 282 acc = true_positive / (true_positive + false_positive + false_negative)
11 E501: line too long (80 > 79 characters) (and 10 similar)
- Line too long (80 > 79 characters):
  6 from gds.common.metrics.metric import Metric, ElementwiseMetric, MultiTaskMetric
- Line too long (89 > 79 characters):
  26 Takes multi-class logits of size (batch_size, ..., n_classes) and returns predictions
- Line too long (112 > 79 characters):
  132 recall = sklearn.metrics.recall_score(y_true, y_pred, average=self.average, labels=torch.unique(y_true))
- Line too long (107 > 79 characters):
  152 score = sklearn.metrics.f1_score(y_true, y_pred, average=self.average, labels=torch.unique(y_true))
- Line too long (105 > 79 characters):
  166 r = pearsonr(y_pred.squeeze().detach().cpu().numpy(), y_true.squeeze().detach().cpu().numpy())[0]
- Line too long (85 > 79 characters):
  178 assert out.dim() > 1, 'MSE loss currently supports Tensors of dimensions > 1'
- Line too long (82 > 79 characters):
  193 """Given a specific model threshold, determine the precision score achieved"""
- Line too long (81 > 79 characters):
  205 return torch.tensor(sklearn.metrics.precision_score(y_true, predictions))
- Line too long (117 > 79 characters):
  256 [self._accuracy(src_boxes["boxes"], pred_boxes, iou_thr) for iou_thr in np.arange(0.5, 0.51, 0.05)]))
- Line too long (83 > 79 characters):
  282 acc = true_positive / (true_positive + false_positive + false_negative)
- Line too long (84 > 79 characters):
  283 return true_positive / (true_positive + false_positive + false_negative)