diff --git a/8_Solow.ipynb b/8_Solow.ipynb
index 361cb3dc..0639394a 100644
--- a/8_Solow.ipynb
+++ b/8_Solow.ipynb
@@ -28,7 +28,7 @@
     "import numpy as np\n",
     "import pandas as pd\n",
     "import py4macro\n",
-    "import statsmodels.formula.api as sm\n",
+    "import statsmodels.formula.api as smf\n",
     "\n",
     "# numpy v1の表示を使用\n",
     "np.set_printoptions(legacy='1.25')\n",
@@ -40,7 +40,8 @@
   {
    "cell_type": "markdown",
    "metadata": {
-    "heading_collapsed": true
+    "heading_collapsed": true,
+    "jp-MarkdownHeadingCollapsed": true
    },
    "source": [
     "## はじめに"
@@ -1431,7 +1432,7 @@
     "for var in ky_ratio.columns[-3:]:              # 1\n",
     "    \n",
     "    df_temp = ky_ratio.copy()                  # 2\n",
-    "    res = sm.ols(f'ky_ratio ~ {var}',          # 3\n",
+    "    res = smf.ols(f'ky_ratio ~ {var}',          # 3\n",
     "                 data=df_temp).fit()           # 4\n",
     "    bhat = res.params.iloc[1]                  # 5\n",
     "    pval = res.pvalues.iloc[1]                 # 6\n",
diff --git a/9_Convergence.ipynb b/9_Convergence.ipynb
index dd89a237..f19b1197 100644
--- a/9_Convergence.ipynb
+++ b/9_Convergence.ipynb
@@ -31,7 +31,7 @@
     "import numpy as np\n",
     "import pandas as pd\n",
     "import py4macro\n",
-    "import statsmodels.formula.api as sm\n",
+    "import statsmodels.formula.api as smf\n",
     "\n",
     "# numpy v1の表示を使用\n",
     "np.set_printoptions(legacy='1.25')\n",
@@ -1532,7 +1532,7 @@
    "outputs": [],
    "source": [
     "formula_absolute = 'gdp_pc_growth ~ gdp_pc_init_log'\n",
-    "res_absolute = sm.ols(formula_absolute, data=df_convergence).fit()\n",
+    "res_absolute = smf.ols(formula_absolute, data=df_convergence).fit()\n",
     "print(res_absolute.summary().tables[1])"
    ]
   },
@@ -1792,7 +1792,7 @@
     "                                        'depreciation +'\n",
     "                                        'gdp_pc_init_log' )\n",
     "\n",
-    "res_conditional = sm.ols(formula_conditional, data=df_convergence).fit()\n",
+    "res_conditional = smf.ols(formula_conditional, data=df_convergence).fit()\n",
     "\n",
     "print(res_conditional.summary().tables[1])"
    ]
@@ -2084,7 +2084,7 @@
     "for yr in range(1950, 2010):                # 7\n",
     "    \n",
     "    df0 = data_for_regression(yr)           # 8 \n",
-    "    res = sm.ols(formula, data=df0).fit()   # 9\n",
+    "    res = smf.ols(formula, data=df0).fit()   # 9\n",
     "    c = res.params                          # 10\n",
     "    p = res.pvalues                         # 11\n",
     "    \n",
@@ -2316,7 +2316,7 @@
     "for yr in range(1950, 2000):                     # 11\n",
     "    \n",
     "    df0 = data_for_regression(yr)                # 12\n",
-    "    res = sm.ols(formula, data=df0).fit()        # 13\n",
+    "    res = smf.ols(formula, data=df0).fit()        # 13\n",
     "    c = res.params                               # 14\n",
     "    p = res.pvalues                              # 15\n",
     "    hypothesis = ( 'saving_rate=0,'              # 16\n",
@@ -2772,7 +2772,7 @@
     "    for yr in range(init_yr, 2000):\n",
     "\n",
     "        df0 = data_for_regression_group(yr, **kwargs)  # 2\n",
-    "        res = sm.ols(formula, data=df0).fit()\n",
+    "        res = smf.ols(formula, data=df0).fit()\n",
     "        c = res.params\n",
     "        p = res.pvalues\n",
     "        hypothesis = ( 'saving_rate=0,'\n",