input fix

app.py

@@ -222,7 +222,6 @@ def plot_brain_slices(data, learning_stage):
     return fig
 
 def interpret_learning_stage(score):
-    """human-readable learning stage"""
     if score < 0.2:
         return "NOVICE: minimal task familiarity, primarily exploratory behavior"
     elif score < 0.4:
@@ -235,23 +234,18 @@ def interpret_learning_stage(score):
         return "EXPERT: automated processing, highly optimized performance"
 
 def plot_results(data, region_acts, temporal_pattern, learning_stage):
-    # 16:9 aspect for modern displays
     fig = plt.figure(figsize=(16, 9))
     
-    # main brain plot: 60% height
     gs = gridspec.GridSpec(2, 2, height_ratios=[6, 4])
     
-    # brain visualization (now larger)
     ax1 = plt.subplot(gs[0, :])
     mean_activation = data.mean(axis=0)
     slice_idx = mean_activation.shape[-1]//2
     brain_slice = mean_activation[...,slice_idx]
     
-    # find activation peaks
     peak_coords = np.unravel_index(np.argmax(brain_slice), brain_slice.shape)
     peak_val = brain_slice[peak_coords]
     
-    # enhanced brain viz
     im = ax1.imshow(brain_slice.T, cmap='hot')
     plt.colorbar(im, ax=ax1)
     ax1.plot(peak_coords[0], peak_coords[1], 'r*', markersize=15)
@@ -261,7 +255,6 @@ def plot_results(data, region_acts, temporal_pattern, learning_stage):
                   fontsize=12, pad=20)
     ax1.axis('off')
     
-    # region plot (lower left)
     ax2 = plt.subplot(gs[1, 0])
     top_n = 5
     region_ranking = np.argsort(-region_acts.flatten())[:top_n]
@@ -270,7 +263,6 @@ def plot_results(data, region_acts, temporal_pattern, learning_stage):
     ax2.set_title('regional activity profile\n' + 
                   'top regions: ' + ', '.join(f'{r}' for r in region_ranking))
     
-    # temporal dynamics (lower right)
     ax3 = plt.subplot(gs[1, 1])
     ax3.plot(temporal_pattern.squeeze(), 'k-', linewidth=2)
     ax3.set_title('temporal evolution')
@@ -286,23 +278,19 @@ def process_fmri(file_obj):
         img = nib.load(file_obj.name)
         data = img.get_fdata(dtype=np.float32)
         
-        # shape validation + expansion
         if data.ndim == 3:
-            data = data[None,...]  # [H,W,D] -> [T,H,W,D]
+            data = data[None,...]  
         elif data.ndim != 4:
             return f"error: volume must be 3D/4D, got {data.ndim}D", None
             
-        # validate dims
         t,h,w,d = data.shape
         if t < 1 or h < 16 or w < 16 or d < 8:
             return f"error: invalid dims {data.shape}, min: [1,16,16,8]", None
         if t > 1000 or h > 256 or w > 256 or d > 256:
             return f"error: dims too large {data.shape}, max: [1000,256,256,256]", None
         
-        # reshape for batch
-        data = data.reshape(1, t, h, w, d)  # [B,T,H,W,D]
+        data = data.reshape(1, t, h, w, d)  
         
-        # normalize + preprocess
         data = preprocess_volume(data)
         device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')
         
@@ -312,7 +300,7 @@ def process_fmri(file_obj):
         for stage in ['full', 'region', 'temporal']:
             try:
                 model = SequentialBrainViT(Config())
-                model._init_weights()  # critical: init before load
+                model._init_weights()  
                 ckpt = torch.load(f'best_{stage}.pt', map_location=device)
                 missing = model.load_state_dict(ckpt['model'], strict=False)
                 if missing:
@@ -328,7 +316,7 @@ def process_fmri(file_obj):
                     }
                     
                     fig = plot_results(
-                        data[0].cpu().numpy(),  # drop batch
+                        data[0].cpu().numpy(), 
                         results[stage]['region_activation'],
                         results[stage]['temporal_pattern'],
                         results[stage]['learning_stage']
@@ -354,11 +342,8 @@ def process_fmri(file_obj):
 
 iface = gr.Interface(
     fn=process_fmri,
-    inputs=gr.File(
-        label="fMRI Data Input",
-        file_types=[".nii", ".nii.gz"],
-        file_count="single"
-    ),
+        inputs=gr.File(label="fMRI data input (.nii/.nii.gz)"),
+
     outputs=[
         gr.Textbox(
             label="Analysis Results", 
@@ -386,8 +371,6 @@ iface = gr.Interface(
     - brain map: warmer colors = higher activation
     - regional profile: shows activity across 116 brain regions
     - temporal pattern: activation changes over time
-    
-    model trained on weatherprediction task fMRI data from openneuro ds000052
     """,
     theme="default",
     examples=[],