不支持 numpy 数组类型？

Question

我正在尝试复制一个我能够遵循并运行教程的模型，但这次使用的是我自己的数据。我能够将自己的 MRI 图像转换为 numpy 数组，其维度与教程数据的数组相同。我尝试用我自己的数组替换教程中的 numpy 数组，并为正常或异常（大小写，不是大小写）编写我自己的虚构 csv 文件。但是，当我运行它时，我得到：(Pytorch) C:\Users\GlaDOS\PythonProjects\dicomnpy>python train.py  -t acl -p sagittal --epochs=10 --prefix_name hueTraceback (most recent call last):  File "train.py", line 277, in <module>    run(args)  File "train.py", line 214, in run    mrnet, train_loader, epoch, num_epochs, optimizer, writer, current_lr, log_every)  File "train.py", line 34, in train_model    for i, (image, label, weight) in enumerate(train_loader):  File "C:\Users\GlaDOS\anaconda3\envs\Pytorch\lib\site-packages\torch\utils\data\dataloader.py", line 345, in __next__    data = self._next_data()  File "C:\Users\GlaDOS\anaconda3\envs\Pytorch\lib\site-packages\torch\utils\data\dataloader.py", line 385, in _next_data    data = self._dataset_fetcher.fetch(index)  # may raise StopIteration  File "C:\Users\GlaDOS\anaconda3\envs\Pytorch\lib\site-packages\torch\utils\data\_utils\fetch.py", line 44, in fetch    data = [self.dataset[idx] for idx in possibly_batched_index]  File "C:\Users\GlaDOS\anaconda3\envs\Pytorch\lib\site-packages\torch\utils\data\_utils\fetch.py", line 44, in <listcomp>    data = [self.dataset[idx] for idx in possibly_batched_index]  File "C:\Users\GlaDOS\PythonProjects\dicomnpy\dataloader.py", line 56, in __getitem__    array = self.transform(array)  File "c:\users\glados\src\torchsample\torchsample\transforms\tensor_transforms.py", line 32, in __call__    inputs = transform(*inputs)  File "C:\Users\GlaDOS\anaconda3\envs\Pytorch\lib\site-packages\torchvision\transforms\transforms.py", line 313, in __call__    return self.lambd(img)  File "train.py", line 167, in <lambda>    transforms.Lambda(lambda x: torch.Tensor(x)),现在我想知道这个错误是否意味着我没有以某种方式将我的 MRI 转换为“正确的”numpy 数组类型？如果是这样，我该如何将它们更改为正确的类型？

Answer 1

您可以重新定义变量astype

your_array = your_array.astype(np.uint16)