example usage

.gitattributes

@@ -48,3 +48,16 @@ v0.3.5.pth filter=lfs diff=lfs merge=lfs -text
 v0.3.6.pth filter=lfs diff=lfs merge=lfs -text
 v0.3.4.pth filter=lfs diff=lfs merge=lfs -text
 v0.3.7.pth filter=lfs diff=lfs merge=lfs -text
+README_files/README_9_0.jpg filter=lfs diff=lfs merge=lfs -text
+README_files/README_9_1.jpg filter=lfs diff=lfs merge=lfs -text
+README_files/README_9_2.jpg filter=lfs diff=lfs merge=lfs -text
+README_files/README_9_3.jpg filter=lfs diff=lfs merge=lfs -text
+README_files/README_4_0.jpg filter=lfs diff=lfs merge=lfs -text
+README_files/README_5_0.jpg filter=lfs diff=lfs merge=lfs -text
+README_files/README_9_1.png filter=lfs diff=lfs merge=lfs -text
+README_files/README_9_2.png filter=lfs diff=lfs merge=lfs -text
+README_files/README_9_3.png filter=lfs diff=lfs merge=lfs -text
+README_files/README_4_0.png filter=lfs diff=lfs merge=lfs -text
+README_files/README_5_0.png filter=lfs diff=lfs merge=lfs -text
+README_files/README_8_0.png filter=lfs diff=lfs merge=lfs -text
+README_files/README_9_0.png filter=lfs diff=lfs merge=lfs -text

README.md

@@ -0,0 +1,142 @@
+```python
+!pip install walloc PyWavelets pytorch-wavelets
+```
+
+
+```python
+!wget "https://r0k.us/graphics/kodak/kodak/kodim05.png"
+```
+
+
+```python
+import torch
+import matplotlib.pyplot as plt
+from PIL import Image
+from IPython.display import display
+from torchvision.transforms import ToPILImage, PILToTensor
+from walloc.walloc import Walloc
+class Args: pass
+```
+
+
+```python
+device = "cpu"
+checkpoint = torch.load("v0.5.1.pth",map_location="cpu")
+args = checkpoint['args']
+codec = Walloc(
+    channels = args.channels,
+    J = args.J,
+    N = args.N,
+    latent_dim = args.latent_dim,
+    latent_bits = 6
+)
+codec.load_state_dict(checkpoint['model_state_dict'])
+codec = codec.to(device)
+```
+
+
+```python
+img = Image.open("kodim05.png")
+img
+```
+
+
+
+
+    
+![png](README_files/README_4_0.png)
+    
+
+
+
+
+```python
+with torch.no_grad():
+    codec.eval()
+    x = PILToTensor()(img).to(torch.float)
+    x = (x/255 - 0.5).unsqueeze(0).to(device)
+    x_hat, _, _ = codec(x)
+ToPILImage()(x_hat[0]+0.5)
+```
+
+
+
+
+    
+![png](README_files/README_5_0.png)
+    
+
+
+
+
+```python
+with torch.no_grad():
+    codec.eval()
+    X = codec.wavelet_analysis(x,J=codec.J)
+    Y = codec.encoder(X)
+    X_hat = codec.decoder(Y)
+    x_hat = codec.wavelet_synthesis(X_hat,J=codec.J)
+
+print(f"dimensionality reduction: {x.numel()/Y.numel()}x")
+```
+
+    dimensionality reduction: 16.0x
+
+
+
+```python
+Y.unique()
+```
+
+
+
+
+    tensor([-31., -30., -29., -28., -27., -26., -25., -24., -23., -22., -21., -20.,
+            -19., -18., -17., -16., -15., -14., -13., -12., -11., -10.,  -9.,  -8.,
+             -7.,  -6.,  -5.,  -4.,  -3.,  -2.,  -1.,   0.,   1.,   2.,   3.,   4.,
+              5.,   6.,   7.,   8.,   9.,  10.,  11.,  12.,  13.,  14.,  15.,  16.,
+             17.,  18.,  19.,  20.,  21.,  22.,  23.,  24.,  25.,  26.,  27.,  28.,
+             29.,  30.,  31.])
+
+
+
+
+```python
+plt.hist(Y.flatten().numpy(),range=(-32,32),bins=65);
+```
+
+
+    
+![png](README_files/README_8_0.png)
+    
+
+
+
+```python
+for latent in Y[0,:4,:,:]:
+    display(ToPILImage()(0.5*latent/codec.latent_max + 0.5).resize((192,128),resample=Image.Resampling.NEAREST))
+```
+
+
+    
+![png](README_files/README_9_0.png)
+    
+
+
+
+    
+![png](README_files/README_9_1.png)
+    
+
+
+
+    
+![png](README_files/README_9_2.png)
+    
+
+
+
+    
+![png](README_files/README_9_3.png)
+    
+