Update app.py

app.py

@@ -6,6 +6,12 @@ from pathlib import Path
 from huggingface_hub import HfApi, Repository
 import matplotlib.pyplot as plt
 
+# Set a clean, sans-serif default font
+plt.rcParams.update({
+    "font.family": "sans-serif",
+    "font.size": 10,
+})
+
 # Global DataFrame
 df = pd.DataFrame()
 
@@ -13,23 +19,34 @@ def upload_csv(file):
     global df
     df = pd.read_csv(file.name)
     if "text" not in df.columns or "label" not in df.columns:
-        return gr.update(visible=False), "❌ CSV must contain ‘text’ and ‘label’ columns."
+        return (
+            gr.update(visible=False),  # hide table
+            "❌ CSV must contain `text` and `label` columns.",
+            gr.update(visible=False),  # hide Save
+            gr.update(visible=False),  # hide Download CSV
+            gr.update(visible=False),  # hide Visualize
+            gr.update(visible=False),  # hide Push Accordion
+        )
     df["label"] = df["label"].fillna("")
     return (
         gr.update(value=df[["text","label"]], visible=True),
-        "✅ File uploaded — you can now edit labels."
+        "✅ File uploaded — you can now annotate.",
+        gr.update(visible=True),
+        gr.update(visible=True),
+        gr.update(visible=True),
+        gr.update(visible=True),
     )
 
-def save_changes(edited_table):
+def save_changes(edited):
     global df
-    df = pd.DataFrame(edited_table, columns=["text","label"])
+    df = pd.DataFrame(edited, columns=["text","label"])
     return "💾 Changes saved."
 
 def download_csv():
     global df
-    out_path = "annotated_data.csv"
-    df.to_csv(out_path, index=False)
-    return out_path
+    path = "annotated_data.csv"
+    df.to_csv(path, index=False)
+    return path
 
 def create_distribution_figure(df_input):
     counts = df_input["label"].value_counts().sort_values(ascending=False)
@@ -37,44 +54,50 @@ def create_distribution_figure(df_input):
     values = counts.values.tolist()
     
     fig, (ax_table, ax_bar) = plt.subplots(
-        nrows=1, ncols=2,
-        gridspec_kw={"width_ratios": [1, 2]},
-        figsize=(8, max(2, len(labels) * 0.3))
+        ncols=2,
+        gridspec_kw={"width_ratios": [1,2]},
+        figsize=(8, max(2, len(labels)*0.4)),
+        tight_layout=True
     )
     
     # Table
     ax_table.axis("off")
     table_data = [[lab, cnt] for lab, cnt in zip(labels, values)]
-    tbl = ax_table.table(cellText=table_data, colLabels=["Label","Count"], loc="center")
+    tbl = ax_table.table(
+        cellText=table_data,
+        colLabels=["Label", "Count"],
+        cellLoc="center",
+        loc="center"
+    )
     tbl.auto_set_font_size(False)
     tbl.set_fontsize(10)
-    tbl.scale(1, 1.5)
+    tbl.scale(1, 1.2)
     
     # Bar chart
-    ax_bar.barh(labels, values)
+    ax_bar.barh(labels, values, color="#222222")
     ax_bar.invert_yaxis()
     ax_bar.set_xlabel("Count")
     ax_bar.set_ylabel("")
     
-    plt.tight_layout()
     return fig
 
 def visualize_and_download_chart():
     global df
     fig = create_distribution_figure(df)
-    chart_path = "label_distribution.png"
-    fig.savefig(chart_path, dpi=150)
-    return fig, chart_path
+    out_path = "label_distribution.png"
+    fig.savefig(out_path, dpi=150, bbox_inches="tight")
+    return fig, out_path
 
-def push_to_hub(repo_name: str, hf_token: str) -> str:
+def push_to_hub(repo_name, hf_token):
     global df
     try:
         api = HfApi()
-        api.create_repo(repo_id=repo_name, token=hf_token, repo_type="dataset", exist_ok=True)
+        api.create_repo(repo_id=repo_name, token=hf_token,
+                        repo_type="dataset", exist_ok=True)
 
-        local_dir = Path(f"./{repo_name.replace('/', '_')}")
+        local_dir = Path(f"./{repo_name.replace('/','_')}")
         if local_dir.exists():
-            for child in local_dir.iterdir(): child.unlink()
+            for f in local_dir.iterdir(): f.unlink()
             local_dir.rmdir()
 
         repo = Repository(
@@ -83,9 +106,7 @@ def push_to_hub(repo_name: str, hf_token: str) -> str:
             repo_type="dataset",
             use_auth_token=hf_token
         )
-
-        csv_path = local_dir / "data.csv"
-        df.to_csv(csv_path, index=False)
+        df.to_csv(local_dir/"data.csv", index=False)
         repo.push_to_hub(commit_message="📑 Update annotated data")
         return f"🚀 Pushed to https://huggingface.co/datasets/{repo_name}"
     except Exception as e:
@@ -93,37 +114,45 @@ def push_to_hub(repo_name: str, hf_token: str) -> str:
 
 with gr.Blocks(theme=gr.themes.Default()) as app:
     gr.Markdown("## 🏷️ Label It! Text Annotation Tool")
-    gr.Markdown("Upload a `.csv` with **text** + **label** columns, annotate in-place, then export, visualize, or publish.")
-
-    with gr.Row():
-        file_input = gr.File(label="📁 Upload CSV", file_types=[".csv"])
-        upload_btn = gr.Button("Upload")
-
-    df_table = gr.Dataframe(headers=["text","label"], label="📝 Editable Table",
-                            interactive=True, visible=False)
-    status = gr.Textbox(label="Status", interactive=False)
-
-    with gr.Row():
-        save_btn        = gr.Button("💾 Save")
-        download_btn    = gr.Button("⬇️ Download CSV")
-        download_out    = gr.File(label="📥 Downloaded File")
-
+    gr.Markdown("**Step 1:** Upload a `.csv` with **text** + **label** columns.")
+    
+    # Step 1: Upload only
     with gr.Row():
-        visualize_btn   = gr.Button("📊 Visualize Distribution")
-        chart_plot      = gr.Plot(label="Label Distribution")
-        download_chart  = gr.File(label="📥 Download Chart")
-
-    with gr.Accordion("📦 Push to Hugging Face Hub", open=False):
-        repo_input  = gr.Textbox(label="Repo (username/dataset-name)")
-        token_input = gr.Textbox(label="🔑 HF Token", type="password")
-        push_btn    = gr.Button("🚀 Push")
-        push_status = gr.Textbox(label="Push Status", interactive=False)
-
-    # Bind events
-    upload_btn.click(upload_csv, inputs=file_input, outputs=[df_table, status])
+        upload_file = gr.File(label="📁 Upload CSV", file_types=[".csv"])
+        upload_btn  = gr.Button("Upload")
+
+    # Hidden until upload
+    df_table      = gr.Dataframe(headers=["text","label"], visible=False, interactive=True)
+    status        = gr.Textbox(label="Status")
+    save_btn      = gr.Button("💾 Save", visible=False)
+    download_btn  = gr.Button("⬇️ Download CSV", visible=False)
+    download_csv_out = gr.File(label="📥 Download CSV", visible=False)
+
+    visualize_btn = gr.Button("📊 Visualize Distribution", visible=False)
+    chart_plot    = gr.Plot(label="Label Distribution", visible=False)
+    download_chart= gr.File(label="📥 Download Chart", visible=False)
+
+    # Push accordion hidden
+    with gr.Accordion("📦 Push to Hugging Face Hub", open=False, visible=False) as push_acc:
+        repo_in  = gr.Textbox(label="Repo (username/dataset-name)")
+        token_in = gr.Textbox(label="🔑 HF Token", type="password")
+        push_btn = gr.Button("🚀 Push")
+        push_status = gr.Textbox(label="Push Status")
+
+    # Bind actions
+    upload_btn.click(
+        upload_csv,
+        inputs=upload_file,
+        outputs=[df_table, status,
+                 save_btn, download_btn, visualize_btn, push_acc]
+    )
     save_btn.click(save_changes, inputs=df_table, outputs=status)
-    download_btn.click(download_csv, outputs=download_out)
-    visualize_btn.click(visualize_and_download_chart, outputs=[chart_plot, download_chart])
-    push_btn.click(push_to_hub, inputs=[repo_input, token_input], outputs=push_status)
+    download_btn.click(download_csv, outputs=download_csv_out)
+    visualize_btn.click(visualize_and_download_chart,
+                        outputs=[chart_plot, download_chart])
+    push_btn.click(push_to_hub, inputs=[repo_in, token_in], outputs=push_status)
+
+    # Step 2 instruction
+    gr.Markdown("**Step 2:** Edit labels, then Save, Visualize or Publish.")
 
 app.launch()