Initial Rotate Function

function.py

@@ -3,6 +3,8 @@
 import numpy as np
 import scipy.ndimage as snd
 from torch.autograd import Variable
+from torchvision.transforms import ToPILImage, ToTensor
+import torchvision.transforms.functional as PIL
 from dataset import VolumeDataset, BlockDataset
 from torch.utils.data import DataLoader
 from model import UNet2d
@@ -22,7 +24,27 @@ def write_nifti(data, aff, shape, out_path):
     img=nib.Nifti1Image(data, aff)
     img.to_filename(out_path)
 
-def rotate_volume(vol): pass
+def rotate_volume(vol):
+    tp_trans=ToPILImage()
+    tt_trans=ToTensor()
+
+    angle=np.array([1, 1, 1])
+    for i in range(3):
+        if i==0:
+            old_vol=vol
+        dim=old_vol.shape[i]
+        for j in range(dim):
+			if i==0:
+            	one_slice=old_vol[j, :, :]
+			elif i==1:
+            	one_slice=old_vol[:, j, :]
+			else: # i==2
+            	one_slice=old_vol[:, :, j]
+			one_slice_pil=tp_trans(one_slice)
+            one_slice_pil=PIL.rotate(one_slice_pil, angle[i], 
+				resample=PIL.Image.BILINEAR, expand=True)
+			one_slice=tt_trans(one_slice_pil)
+			if j==0: pass # Create New Vol
 
 def estimate_dice(gt_msk, prt_msk):
     intersection=gt_msk*prt_msk