Update app.py
Browse files
app.py
CHANGED
@@ -9,7 +9,8 @@ def compare_manifold_learning(methods, n_samples, n_neighbors, n_components, per
|
|
9 |
S_points, S_color = datasets.make_s_curve(n_samples, random_state=0)
|
10 |
transformed_data = []
|
11 |
|
12 |
-
|
|
|
13 |
manifold_method = {
|
14 |
"Locally Linear Embeddings Standard": manifold.LocallyLinearEmbedding(method="standard", n_neighbors=n_neighbors, n_components=n_components, eigen_solver="auto", random_state=0),
|
15 |
"Locally Linear Embeddings LTSA": manifold.LocallyLinearEmbedding(method="ltsa", n_neighbors=n_neighbors, n_components=n_components, eigen_solver="auto", random_state=0),
|
@@ -22,17 +23,42 @@ def compare_manifold_learning(methods, n_samples, n_neighbors, n_components, per
|
|
22 |
}[method]
|
23 |
S_transformed = manifold_method.fit_transform(S_points)
|
24 |
transformed_data.append(S_transformed)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
25 |
|
26 |
fig, axs = plt.subplots(1, len(transformed_data), figsize=(6 * len(transformed_data), 6))
|
27 |
fig.suptitle("Manifold Learning Comparison", fontsize=16)
|
28 |
|
29 |
-
|
|
|
|
|
|
|
30 |
ax.scatter(data[:, 0], data[:, 1], c=S_color, cmap=plt.cm.Spectral)
|
31 |
ax.set_title(f"Method: {method}")
|
32 |
ax.axis("tight")
|
33 |
ax.axis("off")
|
34 |
ax.xaxis.set_major_locator(ticker.NullLocator())
|
35 |
ax.yaxis.set_major_locator(ticker.NullLocator())
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
36 |
|
37 |
plt.tight_layout()
|
38 |
plt.savefig("plot.png")
|
@@ -41,14 +67,14 @@ def compare_manifold_learning(methods, n_samples, n_neighbors, n_components, per
|
|
41 |
return "plot.png"
|
42 |
|
43 |
method_options = [
|
44 |
-
"
|
45 |
-
"
|
46 |
-
"
|
47 |
-
"
|
48 |
"Isomap",
|
49 |
-
"
|
50 |
"Spectral Embedding",
|
51 |
-
"
|
52 |
]
|
53 |
|
54 |
inputs = [
|
|
|
9 |
S_points, S_color = datasets.make_s_curve(n_samples, random_state=0)
|
10 |
transformed_data = []
|
11 |
|
12 |
+
if len(methods) == 1:
|
13 |
+
method = methods[0]
|
14 |
manifold_method = {
|
15 |
"Locally Linear Embeddings Standard": manifold.LocallyLinearEmbedding(method="standard", n_neighbors=n_neighbors, n_components=n_components, eigen_solver="auto", random_state=0),
|
16 |
"Locally Linear Embeddings LTSA": manifold.LocallyLinearEmbedding(method="ltsa", n_neighbors=n_neighbors, n_components=n_components, eigen_solver="auto", random_state=0),
|
|
|
23 |
}[method]
|
24 |
S_transformed = manifold_method.fit_transform(S_points)
|
25 |
transformed_data.append(S_transformed)
|
26 |
+
else:
|
27 |
+
for method in methods:
|
28 |
+
manifold_method = {
|
29 |
+
"Locally Linear Embeddings Standard": manifold.LocallyLinearEmbedding(method="standard", n_neighbors=n_neighbors, n_components=n_components, eigen_solver="auto", random_state=0),
|
30 |
+
"Locally Linear Embeddings LTSA": manifold.LocallyLinearEmbedding(method="ltsa", n_neighbors=n_neighbors, n_components=n_components, eigen_solver="auto", random_state=0),
|
31 |
+
"Locally Linear Embeddings Hessian": manifold.LocallyLinearEmbedding(method="hessian", n_neighbors=n_neighbors, n_components=n_components, eigen_solver="auto", random_state=0),
|
32 |
+
"Locally Linear Embeddings Modified": manifold.LocallyLinearEmbedding(method="modified", n_neighbors=n_neighbors, n_components=n_components, eigen_solver="auto", random_state=0),
|
33 |
+
"Isomap": manifold.Isomap(n_neighbors=n_neighbors, n_components=n_components, p=1),
|
34 |
+
"MultiDimensional Scaling": manifold.MDS(n_components=n_components, max_iter=50, n_init=4, random_state=0, normalized_stress=False),
|
35 |
+
"Spectral Embedding": manifold.SpectralEmbedding(n_components=n_components, n_neighbors=n_neighbors),
|
36 |
+
"T-distributed Stochastic Neighbor Embedding": manifold.TSNE(n_components=n_components, perplexity=perplexity, init="random", n_iter=250, random_state=0)
|
37 |
+
}[method]
|
38 |
+
S_transformed = manifold_method.fit_transform(S_points)
|
39 |
+
transformed_data.append(S_transformed)
|
40 |
|
41 |
fig, axs = plt.subplots(1, len(transformed_data), figsize=(6 * len(transformed_data), 6))
|
42 |
fig.suptitle("Manifold Learning Comparison", fontsize=16)
|
43 |
|
44 |
+
if len(methods) == 1:
|
45 |
+
ax = axs
|
46 |
+
method = methods[0]
|
47 |
+
data = transformed_data[0]
|
48 |
ax.scatter(data[:, 0], data[:, 1], c=S_color, cmap=plt.cm.Spectral)
|
49 |
ax.set_title(f"Method: {method}")
|
50 |
ax.axis("tight")
|
51 |
ax.axis("off")
|
52 |
ax.xaxis.set_major_locator(ticker.NullLocator())
|
53 |
ax.yaxis.set_major_locator(ticker.NullLocator())
|
54 |
+
else:
|
55 |
+
for ax, method, data in zip(axs, methods, transformed_data):
|
56 |
+
ax.scatter(data[:, 0], data[:, 1], c=S_color, cmap=plt.cm.Spectral)
|
57 |
+
ax.set_title(f"Method: {method}")
|
58 |
+
ax.axis("tight")
|
59 |
+
ax.axis("off")
|
60 |
+
ax.xaxis.set_major_locator(ticker.NullLocator())
|
61 |
+
ax.yaxis.set_major_locator(ticker.NullLocator())
|
62 |
|
63 |
plt.tight_layout()
|
64 |
plt.savefig("plot.png")
|
|
|
67 |
return "plot.png"
|
68 |
|
69 |
method_options = [
|
70 |
+
"LLE Standard",
|
71 |
+
"LLE LTSA",
|
72 |
+
"LLE Hessian",
|
73 |
+
"LLE Modified",
|
74 |
"Isomap",
|
75 |
+
"MDS",
|
76 |
"Spectral Embedding",
|
77 |
+
"t-SNE"
|
78 |
]
|
79 |
|
80 |
inputs = [
|