Update app.py

app.py

@@ -1,97 +1,101 @@
-import gradio as gr
-import numpy as np
-
-try:
-    # demo close
-    demo.close()
-except:
-    pass
-
-
-# --- LaTeX形式で行列を表現 ---
-def matrix_to_latex(M, name="A"):
-    rows = [" & ".join(map(lambda x: f"{x:.1f}", row)) for row in M]
-    body = " \\\\ ".join(rows)
-    return f"{name} = \\begin{{bmatrix}} {body} \\end{{bmatrix}}"
-
-# --- ランダム行列生成 ---
-def random_matrix(rows, cols, min_val, max_val, step):
-    values = np.arange(min_val, max_val + step / 2, step)
-    return np.random.choice(values, size=(rows, cols))
-
-# --- 問題生成 ---
-def generate_problem(rows_A, cols_A, cols_B, min_val, max_val, step):
-    rows_B = cols_A  # 行列積が可能なように調整
-    A = random_matrix(int(rows_A), int(cols_A), min_val, max_val, step)
-    B = random_matrix(int(rows_B), int(cols_B), min_val, max_val, step)
-
-    md = (
-        "### 以下の行列の積を求めよ：\n\n"
-        + "$$ " + matrix_to_latex(A, "A") + " $$\n\n"
-        + "$$ " + matrix_to_latex(B, "B") + " $$\n\n"
-        + "求めるもの：$$ A \\times B $$"
-    )
-    return (A.tolist(), B.tolist(), md)
-
-# --- 解答生成 ---
-def show_answer(A, B):
-    A = np.array(A); B = np.array(B)
-    C = A @ B
-
-    steps = []
-    for i in range(C.shape[0]):
-        for j in range(C.shape[1]):
-            terms = [f"{A[i,k]:.1f}\\times{B[k,j]:.1f}" for k in range(A.shape[1])]
-            steps.append(f"$c_{{{i+1}{j+1}}} = " + " + ".join(terms) + f" = {C[i,j]:.1f}$")
-
-    md = (
-        "### 【解答】\n\n"
-        + "$$ " + matrix_to_latex(C, "C") + " $$\n\n"
-        + "### 【計算手順】\n\n"
-        + "\n\n".join(steps)
-    )
-    return md
-
-# --- Gradio UI設定 ---
-latex_delims = [
-    {"left": "$$", "right": "$$", "display": True},
-    {"left": "$", "right": "$", "display": False},
-]
-
-with gr.Blocks(title="行列積ランダム出題") as demo:
-    gr.Markdown("## 🎲 ランダム行列積クイズ\n"
-                "行列サイズや数値範囲を調整して、練習用の問題を自動生成できます。")
-
-    with gr.Accordion("⚙️ パラメータ設定", open=False):
-        with gr.Row():
-            rows_A = gr.Slider(2, 4, value=3, step=1, label="Aの行数")
-            cols_A = gr.Slider(2, 4, value=3, step=1, label="Aの列数（＝Bの行数）")
-            cols_B = gr.Slider(2, 4, value=3, step=1, label="Bの列数")
-        with gr.Row():
-            min_val = gr.Number(value=-5.0, label="最小値（例：-5）")
-            max_val = gr.Number(value=5.0, label="最大値（例：5）")
-            step_val = gr.Number(value=0.5, label="刻み幅（例：0.5）")
-
-    A_state = gr.State()
-    B_state = gr.State()
-
-    problem_md = gr.Markdown("", latex_delimiters=latex_delims, label="問題")
-    answer_md  = gr.Markdown("", latex_delimiters=latex_delims, label="解答")
-
-    with gr.Row():
-        gen_btn = gr.Button("🔄 問題を再作成", variant="primary")
-        ans_btn = gr.Button("💡 解答提示", variant="secondary")
-
-    gen_btn.click(
-        fn=generate_problem,
-        inputs=[rows_A, cols_A, cols_B, min_val, max_val, step_val],
-        outputs=[A_state, B_state, problem_md]
-    )
-
-    ans_btn.click(
-        fn=show_answer,
-        inputs=[A_state, B_state],
-        outputs=answer_md
-    )
-
+import gradio as gr
+import numpy as np
+
+try:
+    # demo close
+    demo.close()
+except:
+    pass
+
+import gradio as gr
+import numpy as np
+
+# --- LaTeX形式で行列を表現 ---
+def matrix_to_latex(M, name="A"):
+    rows = [" & ".join(map(lambda x: f"{x:.1f}", row)) for row in M]
+    body = " \\\\ ".join(rows)
+    return f"{name} = \\begin{{bmatrix}} {body} \\end{{bmatrix}}"
+
+# --- ランダム行列生成 ---
+def random_matrix(rows, cols, min_val, max_val, step):
+    values = np.arange(min_val, max_val + step / 2, step)
+    return np.random.choice(values, size=(rows, cols))
+
+# --- 問題生成 ---
+def generate_problem(rows_A, cols_A, cols_B, min_val, max_val, step):
+    rows_B = cols_A  # 行列積が可能なように調整
+    A = random_matrix(int(rows_A), int(cols_A), min_val, max_val, step)
+    B = random_matrix(int(rows_B), int(cols_B), min_val, max_val, step)
+
+    md = (
+        "### 以下の行列の積を求めよ：\n\n"
+        + "$$ " + matrix_to_latex(A, "A") + " $$\n\n"
+        + "$$ " + matrix_to_latex(B, "B") + " $$\n\n"
+        + "求めるもの：$$ A \\times B $$"
+    )
+    # 新しい問題と一緒に解答欄を空にリセット
+    return (A.tolist(), B.tolist(), md, "")
+
+# --- 解答生成 ---
+def show_answer(A, B):
+    A = np.array(A); B = np.array(B)
+    C = A @ B
+
+    steps = []
+    for i in range(C.shape[0]):
+        for j in range(C.shape[1]):
+            terms = [f"{A[i,k]:.1f}\\times{B[k,j]:.1f}" for k in range(A.shape[1])]
+            steps.append(f"$c_{{{i+1}{j+1}}} = " + " + ".join(terms) + f" = {C[i,j]:.1f}$")
+
+    md = (
+        "### 【解答】\n\n"
+        + "$$ " + matrix_to_latex(C, "C") + " $$\n\n"
+        + "### 【計算手順】\n\n"
+        + "\n\n".join(steps)
+    )
+    return md
+
+# --- Gradio UI設定 ---
+latex_delims = [
+    {"left": "$$", "right": "$$", "display": True},
+    {"left": "$", "right": "$", "display": False},
+]
+
+with gr.Blocks(title="行列積ランダム出題") as demo:
+    gr.Markdown("## 🎲 ランダム行列積クイズ\n"
+                "行列サイズや数値範囲を調整して、練習用の問題を自動生成できます。")
+
+    with gr.Accordion("⚙️ パラメータ設定", open=False):
+        with gr.Row():
+            rows_A = gr.Slider(2, 4, value=3, step=1, label="Aの行数")
+            cols_A = gr.Slider(2, 4, value=3, step=1, label="Aの列数（＝Bの行数）")
+            cols_B = gr.Slider(2, 4, value=3, step=1, label="Bの列数")
+        with gr.Row():
+            min_val = gr.Number(value=-5.0, label="最小値（例：-5）")
+            max_val = gr.Number(value=5.0, label="最大値（例：5）")
+            step_val = gr.Number(value=0.5, label="刻み幅（例：0.5）")
+
+    A_state = gr.State()
+    B_state = gr.State()
+
+    problem_md = gr.Markdown("", latex_delimiters=latex_delims, label="問題")
+    answer_md  = gr.Markdown("", latex_delimiters=latex_delims, label="解答")
+
+    with gr.Row():
+        gen_btn = gr.Button("🔄 問題を再作成", variant="primary")
+        ans_btn = gr.Button("💡 解答提示", variant="secondary")
+
+    # 新しい問題作成時に、解答欄もクリア
+    gen_btn.click(
+        fn=generate_problem,
+        inputs=[rows_A, cols_A, cols_B, min_val, max_val, step_val],
+        outputs=[A_state, B_state, problem_md, answer_md]
+    )
+
+    ans_btn.click(
+        fn=show_answer,
+        inputs=[A_state, B_state],
+        outputs=answer_md
+    )
+
 demo.launch()