Merge branch 'main' of github.com:agosztolai/MARBLE into main

MARBLE/geometry.py

@@ -64,7 +64,7 @@ def furthest_point_sampling(x, N=None, spacing=0.0, start_idx=0):
     return perm, lambdas
 
 
-def cluster(x, cluster_typ="meanshift", n_clusters=15, seed=0):
+def cluster(x, cluster_typ="kmeans", n_clusters=15, seed=0):
     """Cluster data.
 
     Args:
@@ -78,7 +78,7 @@ def cluster(x, cluster_typ="meanshift", n_clusters=15, seed=0):
     """
     clusters = {}
     if cluster_typ == "kmeans":
-        kmeans = KMeans(n_clusters=n_clusters, random_state=seed, n_init="auto").fit(x)
+        kmeans = KMeans(n_clusters=n_clusters, random_state=seed).fit(x)
         clusters["n_clusters"] = n_clusters
         clusters["labels"] = kmeans.labels_
         clusters["centroids"] = kmeans.cluster_centers_

README.md

@@ -30,7 +30,7 @@ See full documentation [here](https://agosztolai.github.io/MARBLE/).
 
 ## Installation
 
-The code is tested for CPU and GPU (CUDA) machines running Linux or OSX. Although smaller examples run fast on CPU, for larger datasets, it is highly recommended that you use a GPU machine.
+The code is tested for CPU and GPU (CUDA) machines running Linux, Mac OSX or Windows. Although smaller examples run fast on CPU, for larger datasets, it is highly recommended that you use a GPU machine.
 
 We recommend you install the code in a fresh Anaconda virtual environment, as follows.
 
@@ -76,14 +76,14 @@ We suggest you study at least the example of a [simple vector fields over flat s
 
 Briefly, MARBLE takes two inputs
 
-1. `pos` - a list of `nxd` arrays, each defining a point cloud describing the geometry of a manifold
-2. `x` - a list of `nxD` arrays, defining a signal over the respective manifolds in 1. For dynamical systems, D=d, but our code can also handle signals of other dimensions. Read more about [inputs](#inputs) and [different conditions](#conditions).
+1. `pos` - a list of `nxd` arrays, each defining a cloud of anchor points describing the geometry of a manifold
+2. `x` - a list of `nxD` arrays, defining a vector signal over the respective manifolds in 1. For dynamical systems, D=d, but our code can also handle signals of other dimensions. Read more about [inputs](#inputs) and [different conditions](#conditions).
 
 Using these inputs, you can construct a dataset for MARBLE.
 
 ```
 import MARBLE 
-data = MARBLE.construct_dataset(pos, features=x)
+data = MARBLE.construct_dataset(anchor=pos, vector=x)
 ```
 
 The main attributes are `data.pos` - manifold positions concatenated, `data.x` - manifold signals concatenated and `data.y` - identifiers that tell you which manifold the point belongs to. Read more about [other usedul data attributed](#construct).
@@ -96,7 +96,7 @@ model = MARBLE.net(data)
 model.fit(data)
 ```
 
-By default, MARBLE operated in geometry-aware mode. You can enable the geometry-agnostic mode by changing the initialisation step to
+By default, MARBLE operates in geometry-aware mode. You can enable the geometry-agnostic mode by changing the initialisation step to
 
 ```
 model = MARBLE.net(data, params = {'inner_product_features': True})

examples/macaque_reaching/README.md

@@ -12,7 +12,8 @@ The scripts and notebooks are as follows.
 
 - convert_data.py: convert original spiking and kinematics data to data compatible with MARBLE (both input and generated data are on dataverse)
 - run_marble.py: train MARBLE networks (this takes a while to run - 1-2h/session)
-- analysis.ipynb: create plots of some panels of Figure 3
 
-3. Decoding into kinematics
+3. If you want to simply reproduce the results, run - analysis.ipynb: create plots of some panels of Figure 3
+
+4. Decoding into kinematics
 - decoding.ipynb

examples/toy_examples/vanderpol.ipynb

@@ -184,7 +184,7 @@
     }
    ],
    "source": [
-    "data = MARBLE.construct_dataset(pos, features=vel, stop_crit=0.03)"
+    "data = MARBLE.construct_dataset(anchor=pos, vector=vel, spacing=0.03)"
    ]
   },
   {
@@ -513,7 +513,7 @@
     "mus = np.linspace(-0.1, 0.1, n_mus)\n",
     "pos, vel = get_pos_vel(mus)\n",
     "\n",
-    "data = MARBLE.construct_dataset(pos, features=vel, k=20, stop_crit=0.03)\n",
+    "data = MARBLE.construct_dataset(anchor=pos, vector=vel, k=20, spacing=0.03)\n",
     "model = MARBLE.net(data, params=params)\n",
     "model.fit(data, outdir='model_zoom')"
    ]
@@ -759,7 +759,7 @@
     "mus = np.linspace(-0.1, 0.1, n_mus)\n",
     "pos, vel = get_pos_vel_noise(mus, alpha=0.2)\n",
     "\n",
-    "data = MARBLE.construct_dataset(pos, features=vel, k=20, stop_crit=0.03)\n",
+    "data = MARBLE.construct_dataset(anchor=pos, vector=vel, k=20, spacing=0.03)\n",
     "\n",
     "params['inner_product_features'] = True\n",
     "model = MARBLE.net(data, params=params)\n",