Merge pull request #354 from BindsNET/hananel

Hananel

Author: Hananel-Hazan

bindsnet/analysis/plotting.py

@@ -468,15 +468,15 @@ def plot_voltages(
 
     if time is None:
         for key in voltages.keys():
-            time = (0, voltages[key].size(-1))
+            time = (0, voltages[key].size(0))
             break
 
     if n_neurons is None:
         n_neurons = {}
 
     for key, val in voltages.items():
         if key not in n_neurons.keys():
-            n_neurons[key] = (0, val.size(0))
+            n_neurons[key] = (0, val.size(1))
 
     if not ims:
         fig, axes = plt.subplots(n_subplots, 1, figsize=figsize)
@@ -538,8 +538,8 @@ def plot_voltages(
                             v[1]
                             .cpu()
                             .numpy()[
+                                time[0]: time[1],
                                 n_neurons[v[0]][0] : n_neurons[v[0]][1],
-                                time[0] : time[1],
                             ]
                         )
                     )
@@ -557,8 +557,8 @@ def plot_voltages(
                             v[1]
                             .cpu()
                             .numpy()[
+                                time[0]: time[1],
                                 n_neurons[v[0]][0] : n_neurons[v[0]][1],
-                                time[0] : time[1],
                             ]
                             .T,
                             cmap=cmap,
@@ -621,7 +621,8 @@ def plot_voltages(
                         v[1]
                         .cpu()
                         .numpy()[
-                            n_neurons[v[0]][0] : n_neurons[v[0]][1], time[0] : time[1]
+                            time[0]: time[1],
+                            n_neurons[v[0]][0]: n_neurons[v[0]][1],
                         ]
                     )
                     if thresholds is not None and thresholds[v[0]].size() == torch.Size(
@@ -635,7 +636,8 @@ def plot_voltages(
                         v[1]
                         .cpu()
                         .numpy()[
-                            n_neurons[v[0]][0] : n_neurons[v[0]][1], time[0] : time[1]
+                            time[0]: time[1],
+                            n_neurons[v[0]][0]: n_neurons[v[0]][1],
                         ]
                         .T,
                         cmap=cmap,

examples/mnist/supervised_mnist.py

@@ -118,6 +118,9 @@
 # Sequence of accuracy estimates.
 accuracy = {"all": [], "proportion": []}
 
+# Labels to determine neuron assignments and spike proportions and estimate accuracy
+labels = torch.empty(update_interval)
+
 spikes = {}
 for layer in set(network.layers):
     spikes[layer] = Monitor(network.layers[layer], state_vars=["s"], time=time)
@@ -154,10 +157,10 @@
 
         # Compute network accuracy according to available classification strategies.
         accuracy["all"].append(
-            100 * torch.sum(label.long() == all_activity_pred).item() / update_interval
+            100 * torch.sum(labels.long() == all_activity_pred).item() / update_interval
         )
         accuracy["proportion"].append(
-            100 * torch.sum(label.long() == proportion_pred).item() / update_interval
+            100 * torch.sum(labels.long() == proportion_pred).item() / update_interval
         )
 
         print(
@@ -174,13 +177,16 @@
         )
 
         # Assign labels to excitatory layer neurons.
-        assignments, proportions, rates = assign_labels(spike_record, label, 10, rates)
+        assignments, proportions, rates = assign_labels(spike_record, labels, 10, rates)
+
+    #Add the current label to the list of labels for this update_interval
+    labels[i % update_interval] = label[0]
 
     # Run the network on the input.
     choice = np.random.choice(int(n_neurons / 10), size=n_clamp, replace=False)
     clamp = {"Ae": per_class * label.long() + torch.Tensor(choice).long()}
-    inputs = {"X": image.view(time, 1, 28, 28)}
-    network.run(inputs=inputs, time=time, clamp=clamp)
+    inputs = {"X": image.view(time, 1, 1, 28, 28)}
+    network.run(inpts=inputs, time=time, clamp=clamp)
 
     # Get voltage recording.
     exc_voltages = exc_voltage_monitor.get("v")
@@ -203,7 +209,7 @@
             image.sum(1).view(28, 28), inpt, label=label, axes=inpt_axes, ims=inpt_ims
         )
         spike_ims, spike_axes = plot_spikes(
-            {layer: spikes[layer].get("s") for layer in spikes},
+            {layer: spikes[layer].get("s").view(time, 1, -1) for layer in spikes},
             ims=spike_ims,
             axes=spike_axes,
         )