How to get class labels from predictions

Part 1 (2019)
1

So I was trying to predict on a test set using learn.get_preds(ds_type=DatasetType.Test)
and I got the predictions. But I need the labels of the top k predicted classes. Can anyone tell me how to do that

1 Like
2

In fastai v1.0.27 you can get class to index mapping using learn.data.c2i. You can use torch.topk to get the indexes of the highest k values. Use these to get the class labels.

5 Likes
3

Thanks I didn’t know fastai had such function

4

No problem :slight_smile: We are all learning new things from each other.

5

Getting this error while running learn.get_preds(ds_type=DatasetType.Test) 

IndexError                                Traceback (most recent call last)
<ipython-input-10-ce77ea7f5e1e> in <module>
----> 1 learn.get_preds(ds_type=DatasetType.Test)

/opt/anaconda3/lib/python3.6/site-packages/fastai/basic_train.py in get_preds(self, ds_type, with_loss, n_batch, pbar)
    209         lf = self.loss_func if with_loss else None
    210         return get_preds(self.model, self.dl(ds_type), cb_handler=CallbackHandler(self.callbacks),
--> 211                          activ=_loss_func2activ(self.loss_func), loss_func=lf, n_batch=n_batch, pbar=pbar)
    212 
    213     def pred_batch(self, ds_type:DatasetType=DatasetType.Valid, pbar:Optional[PBar]=None) -> List[Tensor]:

/opt/anaconda3/lib/python3.6/site-packages/fastai/basic_train.py in get_preds(model, dl, pbar, cb_handler, activ, loss_func, n_batch)
     36     "Tuple of predictions and targets, and optional losses (if `loss_func`) using `dl`, max batches `n_batch`."
     37     res = [torch.cat(o).cpu() for o in
---> 38            zip(*validate(model, dl, cb_handler=cb_handler, pbar=pbar, average=False, n_batch=n_batch))]
     39     if loss_func is not None: res.append(calc_loss(res[0], res[1], loss_func))
     40     if activ is not None: res[0] = activ(res[0])

/opt/anaconda3/lib/python3.6/site-packages/fastai/basic_train.py in validate(model, dl, loss_func, cb_handler, pbar, average, n_batch)
     47     with torch.no_grad():
     48         val_losses,nums = [],[]
---> 49         for xb,yb in progress_bar(dl, parent=pbar, leave=(pbar is not None)):
     50             if cb_handler: xb, yb = cb_handler.on_batch_begin(xb, yb, train=False)
     51             val_losses.append(loss_batch(model, xb, yb, loss_func, cb_handler=cb_handler))

/opt/anaconda3/lib/python3.6/site-packages/fastprogress/fastprogress.py in __iter__(self)
     63         self.update(0)
     64         try:
---> 65             for i,o in enumerate(self._gen):
     66                 yield o
     67                 if self.auto_update: self.update(i+1)

/opt/anaconda3/lib/python3.6/site-packages/fastai/basic_data.py in __iter__(self)
     45     def __iter__(self):
     46         "Process and returns items from `DataLoader`."
---> 47         for b in self.dl:
     48             y = b[1][0] if is_listy(b[1]) else b[1]
     49             if not self.skip_size1 or y.size(0) != 1: yield self.proc_batch(b)

/opt/anaconda3/lib/python3.6/site-packages/torch/utils/data/dataloader.py in __next__(self)
    635                 self.reorder_dict[idx] = batch
    636                 continue
--> 637             return self._process_next_batch(batch)
    638 
    639     next = __next__  # Python 2 compatibility

/opt/anaconda3/lib/python3.6/site-packages/torch/utils/data/dataloader.py in _process_next_batch(self, batch)
    656         self._put_indices()
    657         if isinstance(batch, ExceptionWrapper):
--> 658             raise batch.exc_type(batch.exc_msg)
    659         return batch
    660 

IndexError: Traceback (most recent call last):
  File "/opt/anaconda3/lib/python3.6/site-packages/torch/utils/data/dataloader.py", line 138, in _worker_loop
    samples = collate_fn([dataset[i] for i in batch_indices])
  File "/opt/anaconda3/lib/python3.6/site-packages/torch/utils/data/dataloader.py", line 138, in <listcomp>
    samples = collate_fn([dataset[i] for i in batch_indices])
  File "/opt/anaconda3/lib/python3.6/site-packages/fastai/data_block.py", line 415, in __getitem__
    if self.item is None: x,y = self.x[idxs],self.y[idxs]
  File "/opt/anaconda3/lib/python3.6/site-packages/fastai/data_block.py", line 82, in __getitem__
    if isinstance(try_int(idxs), int): return self.get(idxs)
  File "/opt/anaconda3/lib/python3.6/site-packages/fastai/vision/data.py", line 289, in get
    fn = super().get(i)
  File "/opt/anaconda3/lib/python3.6/site-packages/fastai/data_block.py", line 52, in get
    def get(self, i)->Any: return self.items[i]
IndexError: index 0 is out of bounds for axis 0 with size 0
6

Do you know the reason for integer encoding the classes ?
To me the most intuitive encoding would be a binary encoding, such that for each image, the label would be a vector of length len(classes), with 1s wherever the class is labeled positive.
I am simply wondering why the almighty creators of fastAI chose to use dictionaries instead.

7

don’t you think that would be quite memory intensive as compared to encoding them into dictionaries. Think of a situation where there are many classes. The one-hot vector would become quite sparse as compared to a dictionary

1 Like