Upload app.py with huggingface_hub

app.py

@@ -21,7 +21,7 @@ import numpy as np
 import pandas as pd
 from sklearn.manifold import TSNE
 from sklearn.decomposition import PCA
-import plotly.express as px
+import matplotlib.pyplot as plt
 
 # Load data
 def load_data():
@@ -125,26 +125,29 @@ def plot_tsne(tech_filter, snr_filter, mod_filter, mob_filter, representation, c
     x_pad = max(1e-3, (x_max - x_min) * 0.05)
     y_pad = max(1e-3, (y_max - y_min) * 0.05)
 
-    # Plot (force serialization-friendly types)
-    plot_df = filtered_df.copy()
-    plot_df['x'] = plot_df['x'].astype(float)
-    plot_df['y'] = plot_df['y'].astype(float)
-
-    fig = px.scatter(
-        plot_df,
-        x='x',
-        y='y',
-        color=color_by,
-        hover_data=['tech', 'snr', 'mod', 'mob'],
-        title=f"t-SNE of {representation} ({len(filtered_df)} samples)",
-        template="plotly_white",
-        render_mode="auto",
-        color_discrete_sequence=px.colors.qualitative.Set2 + px.colors.qualitative.Set3,
-    )
-    fig.update_layout(legend_title_text=color_by.capitalize(), margin=dict(l=40, r=20, t=60, b=40))
-    fig.update_traces(mode='markers', marker=dict(size=8, opacity=0.9, line=dict(width=0)))
-    fig.update_xaxes(range=[x_min - x_pad, x_max + x_pad], zeroline=False)
-    fig.update_yaxes(range=[y_min - y_pad, y_max + y_pad], zeroline=False)
+    # Plot using matplotlib for maximum reliability in Spaces
+    fig, ax = plt.subplots(figsize=(7, 6))
+    # cycle through a tab20 palette
+    colors = plt.cm.tab20(np.linspace(0, 1, len(filtered_df[color_by].unique())))
+    for c, cls in zip(colors, sorted(filtered_df[color_by].unique())):
+        mask = filtered_df[color_by] == cls
+        ax.scatter(
+            filtered_df.loc[mask, 'x'],
+            filtered_df.loc[mask, 'y'],
+            s=18,
+            alpha=0.8,
+            label=str(cls),
+            color=c,
+            edgecolors='none',
+        )
+    ax.set_xlim(x_min - x_pad, x_max + x_pad)
+    ax.set_ylim(y_min - y_pad, y_max + y_pad)
+    ax.set_xlabel("t-SNE x")
+    ax.set_ylabel("t-SNE y")
+    ax.set_title(f"t-SNE of {representation} ({len(filtered_df)} samples)")
+    ax.grid(True, alpha=0.3)
+    ax.legend(title=color_by, fontsize=9, title_fontsize=10, loc='best')
+    fig.tight_layout()
     
     coord_info = f"x[{x_min:.3f},{x_max:.3f}] y[{y_min:.3f},{y_max:.3f}]"
     trace_info = f"traces: {len(fig.data)}"