diff --git a/_doc/notebooks/onnx_float32_and_64.ipynb b/_doc/notebooks/onnx_float32_and_64.ipynb
index 36e3afd3e..747e2fd09 100644
--- a/_doc/notebooks/onnx_float32_and_64.ipynb
+++ b/_doc/notebooks/onnx_float32_and_64.ipynb
@@ -196,7 +196,7 @@
         {
           "data": {
             "text/plain": [
-              "0.7836975180890742"
+              "0.77265840833068"
             ]
           },
           "execution_count": 4,
@@ -216,10 +216,10 @@
         {
           "data": {
             "text/plain": [
-              "array([-1.02626125e-01,  5.37157286e-02,  4.85401940e-04,  2.29933224e+00,\n",
-              "       -1.55768235e+01,  2.95197696e+00,  1.27833099e-02, -1.50176256e+00,\n",
-              "        3.86793001e-01, -1.46468258e-02, -9.10378501e-01,  9.96266806e-03,\n",
-              "       -6.34832141e-01])"
+              "array([-1.15252395e-01,  4.88204037e-02, -8.41963443e-03,  2.59331874e+00,\n",
+              "       -2.06335208e+01,  4.04718591e+00, -2.11526920e-03, -1.65794672e+00,\n",
+              "        2.84796430e-01, -9.98951697e-03, -9.90364341e-01,  1.03828254e-02,\n",
+              "       -4.68964825e-01])"
             ]
           },
           "execution_count": 5,
@@ -239,7 +239,7 @@
         {
           "data": {
             "text/plain": [
-              "40.84518080004733"
+              "36.835989825553355"
             ]
           },
           "execution_count": 6,
@@ -266,7 +266,7 @@
         {
           "data": {
             "text/plain": [
-              "array([34.02095272,  8.03158922, 22.99227105, 21.87550536,  9.38806914])"
+              "array([23.62794378, 21.20532332, 27.58860225, 18.6063024 , 15.93082815])"
             ]
           },
           "execution_count": 7,
@@ -287,7 +287,7 @@
         {
           "data": {
             "text/plain": [
-              "array([34.02095272,  8.03158922, 22.99227105, 21.87550536,  9.38806914])"
+              "array([23.62794378, 21.20532332, 27.58860225, 18.6063024 , 15.93082815])"
             ]
           },
           "execution_count": 8,
@@ -346,16 +346,16 @@
         {
           "data": {
             "text/html": [
-              "<div id=\"M8b0b2166a67e4e8dbb50bce246e2d25f-cont\"><div id=\"M8b0b2166a67e4e8dbb50bce246e2d25f\" style=\"width:100%;height:100%;\"></div></div>\n",
+              "<div id=\"Mdcee47589f1649d9865d6bae4570a8b9-cont\"><div id=\"Mdcee47589f1649d9865d6bae4570a8b9\" style=\"width:100%;height:100%;\"></div></div>\n",
               "<script>\n",
               "\n",
-              "require(['http://www.xavierdupre.fr/js/vizjs/viz.js'], function() { var svgGraph = Viz(\"digraph{\\n  nodesep=0.05;\\n  ranksep=0.25;\\n  orientation=portrait;\\n\\n  X [shape=box color=red label=\\\"X\\nfloat((127, 13))\\\" fontsize=10];\\n\\n  Y [shape=box color=green label=\\\"Y\\nfloat((127,))\\\" fontsize=10];\\n\\n  Ma_MatMulcst [shape=box label=\\\"Ma_MatMulcst\\nfloat32((13,))\\n[-1.0262612e-01  5.3715728e-02...\\\" fontsize=10];\\n  Ad_Addcst [shape=box label=\\\"Ad_Addcst\\nfloat32((1,))\\n[40.84518]\\\" fontsize=10];\\n\\n  Ma_O0 [shape=box label=\\\"Ma_O0\\\" fontsize=10];\\n  Ma_MatMul [shape=box style=\\\"filled,rounded\\\" color=orange label=\\\"MatMul\\n(Ma_MatMul)\\\" fontsize=10];\\n  X -> Ma_MatMul;\\n  Ma_MatMulcst -> Ma_MatMul;\\n  Ma_MatMul -> Ma_O0;\\n\\n  Ad_Add [shape=box style=\\\"filled,rounded\\\" color=orange label=\\\"Add\\n(Ad_Add)\\\" fontsize=10];\\n  Ma_O0 -> Ad_Add;\\n  Ad_Addcst -> Ad_Add;\\n  Ad_Add -> Y;\\n}\");\n",
-              "document.getElementById('M8b0b2166a67e4e8dbb50bce246e2d25f').innerHTML = svgGraph; });\n",
+              "require(['http://www.xavierdupre.fr/js/vizjs/viz.js'], function() { var svgGraph = Viz(\"digraph{\\n  ranksep=0.25;\\n  orientation=portrait;\\n  nodesep=0.05;\\n\\n  X [shape=box color=red label=\\\"X\\nfloat((0, 13))\\\" fontsize=10];\\n\\n  Y [shape=box color=green label=\\\"Y\\nfloat((0,))\\\" fontsize=10];\\n\\n  Ma_MatMulcst [shape=box label=\\\"Ma_MatMulcst\\nfloat32((13,))\\n[-1.15252398e-01  4.88204025e-02 -8.41963477e-03  ...\\\" fontsize=10];\\n  Ad_Addcst [shape=box label=\\\"Ad_Addcst\\nfloat32((1,))\\n[36.83599]\\\" fontsize=10];\\n\\n  Ma_Y0 [shape=box label=\\\"Ma_Y0\\\" fontsize=10];\\n  Ma_MatMul [shape=box style=\\\"filled,rounded\\\" color=orange label=\\\"MatMul\\n(Ma_MatMul)\\\" fontsize=10];\\n  X -> Ma_MatMul;\\n  Ma_MatMulcst -> Ma_MatMul;\\n  Ma_MatMul -> Ma_Y0;\\n\\n  Ad_Add [shape=box style=\\\"filled,rounded\\\" color=orange label=\\\"Add\\n(Ad_Add)\\\" fontsize=10];\\n  Ma_Y0 -> Ad_Add;\\n  Ad_Addcst -> Ad_Add;\\n  Ad_Add -> Y;\\n}\");\n",
+              "document.getElementById('Mdcee47589f1649d9865d6bae4570a8b9').innerHTML = svgGraph; });\n",
               "\n",
               "</script>"
             ],
             "text/plain": [
-              "<jyquickhelper.jspy.render_nb_js_dot.RenderJsDot at 0x2d1f3f9feb8>"
+              "<jyquickhelper.jspy.render_nb_js_dot.RenderJsDot at 0x2c04cbd7e48>"
             ]
           },
           "execution_count": 11,
@@ -392,7 +392,7 @@
         {
           "data": {
             "text/plain": [
-              "array([34.020954 ,  8.0315895, 22.99227  , 21.875505 ,  9.388069 ],\n",
+              "array([23.627945, 21.205326, 27.588604, 18.606304, 15.930832],\n",
               "      dtype=float32)"
             ]
           },
@@ -423,7 +423,7 @@
         {
           "data": {
             "text/plain": [
-              "array([34.020954 ,  8.0315895, 22.992271 , 21.875505 ,  9.388065 ],\n",
+              "array([23.627945, 21.205326, 27.588604, 18.606304, 15.930832],\n",
               "      dtype=float32)"
             ]
           },
@@ -475,7 +475,7 @@
         {
           "data": {
             "text/plain": [
-              "array([34.02095272,  8.03158922, 22.99227105, 21.87550536,  9.38806914])"
+              "array([23.62794378, 21.20532332, 27.58860225, 18.6063024 , 15.93082815])"
             ]
           },
           "execution_count": 15,
@@ -500,15 +500,7 @@
       "cell_type": "code",
       "execution_count": 15,
       "metadata": {},
-      "outputs": [
-        {
-          "name": "stdout",
-          "output_type": "stream",
-          "text": [
-            "[ONNXRuntimeError] : 9 : NOT_IMPLEMENTED : Could not find an implementation for the node Ad_Add:Add(7)\n"
-          ]
-        }
-      ],
+      "outputs": [],
       "source": [
         "try:\n",
         "    oinf = OnnxInference(onnx_model64, runtime=\"onnxruntime1\")\n",
@@ -561,10 +553,11 @@
         {
           "data": {
             "text/plain": [
-              "array([33.125,  6.625, 23.5  , 22.625, 10.25 , 19.75 , 13.75 , 17.25 ,\n",
-              "       31.   , 16.   , 25.625, 18.375, 33.625, 42.625, 35.75 , 21.   ,\n",
-              "       19.   , 22.625, 24.75 , 15.   , 31.   ,  7.625, 13.5  ,  4.   ,\n",
-              "       25.25 ], dtype=float32)"
+              "array([24.375  , 20.625  , 28.5    , 18.3125 , 18.15625, 15.5    ,\n",
+              "       40.40625, 21.0625 , 17.5    , 16.25   , 18.625  , 23.5    ,\n",
+              "       20.875  , 16.8125 , 18.75   , 25.375  , 21.84375, 27.75   ,\n",
+              "       14.75   , 31.65625, 20.40625, 21.9375 , 26.5    , 13.53125,\n",
+              "       34.15625], dtype=float32)"
             ]
           },
           "execution_count": 18,
@@ -588,11 +581,11 @@
         {
           "data": {
             "text/plain": [
-              "array([33.16219135,  6.80583637, 23.8054958 , 23.02496476, 10.0123476 ,\n",
-              "       19.74633171, 14.22049743, 17.50913404, 31.23975419, 15.5676655 ,\n",
-              "       25.91874293, 19.14989035, 33.89588065, 43.37730604, 35.97438905,\n",
-              "       21.04844868, 19.1410149 , 22.71168399, 25.04743772, 14.6211457 ,\n",
-              "       31.13142217,  7.47344522, 14.59039259,  4.16909152, 24.71047251])"
+              "array([24.68275083, 20.17216288, 28.4280004 , 18.16998445, 18.26424628,\n",
+              "       16.07043206, 40.59617799, 20.4861235 , 17.96905057, 16.24325372,\n",
+              "       19.01950901, 23.65422543, 20.54061589, 16.83765505, 18.85620332,\n",
+              "       25.66149306, 21.4924616 , 27.50628927, 14.67275618, 31.56870577,\n",
+              "       20.16713281, 21.42304151, 26.96992261, 13.81531227, 34.04079153])"
             ]
           },
           "execution_count": 19,
@@ -622,7 +615,7 @@
         {
           "data": {
             "text/plain": [
-              "array([0.73869556, 0.7735361 , 0.8158365 , 0.90728579, 1.63133946])"
+              "array([0.47186822, 0.51445849, 0.52629847, 0.5763765 , 0.70492867])"
             ]
           },
           "execution_count": 20,
@@ -658,7 +651,7 @@
         {
           "data": {
             "text/plain": [
-              "array([0.73869556, 0.7735361 , 0.8158365 , 0.90728579, 1.63133946])"
+              "array([0.47186822, 0.51445849, 0.52629847, 0.5763765 , 0.70492867])"
             ]
           },
           "execution_count": 22,
@@ -705,16 +698,16 @@
         {
           "data": {
             "text/html": [
-              "<div id=\"M1c20a2ae95eb48c58d8ebc34dc0793da-cont\"><div id=\"M1c20a2ae95eb48c58d8ebc34dc0793da\" style=\"width:100%;height:100%;\"></div></div>\n",
+              "<div id=\"Md6d6f79f149e4869a5c902cd901836e4-cont\"><div id=\"Md6d6f79f149e4869a5c902cd901836e4\" style=\"width:100%;height:100%;\"></div></div>\n",
               "<script>\n",
               "\n",
-              "require(['http://www.xavierdupre.fr/js/vizjs/viz.js'], function() { var svgGraph = Viz(\"digraph{\\n  nodesep=0.05;\\n  ranksep=0.25;\\n  orientation=portrait;\\n\\n  X [shape=box color=red label=\\\"X\\ndouble((1, 13))\\\" fontsize=10];\\n\\n  GPmean [shape=box color=green label=\\\"GPmean\\nfloat((1, 1))\\\" fontsize=10];\\n\\n  kgpd_MatMulcst [shape=box label=\\\"kgpd_MatMulcst\\nfloat64((13, 379))\\n[[1.7120e-01 7.8960e-02 8.3080...\\\" fontsize=10];\\n  kgpd_Addcst [shape=box label=\\\"kgpd_Addcst\\nfloat64((1,))\\n[1073.54778074]\\\" fontsize=10];\\n  gpr_MatMulcst [shape=box label=\\\"gpr_MatMulcst\\nfloat64((379,))\\n[ 2.75920907e-01 -1.82627201e-...\\\" fontsize=10];\\n  gpr_Addcst [shape=box label=\\\"gpr_Addcst\\nfloat64((1, 1))\\n[[0.]]\\\" fontsize=10];\\n\\n  kgpd_O02 [shape=box label=\\\"kgpd_O02\\\" fontsize=10];\\n  kgpd_MatMul [shape=box style=\\\"filled,rounded\\\" color=orange label=\\\"MatMul\\n(kgpd_MatMul)\\\" fontsize=10];\\n  X -> kgpd_MatMul;\\n  kgpd_MatMulcst -> kgpd_MatMul;\\n  kgpd_MatMul -> kgpd_O02;\\n\\n  kgpd_O01 [shape=box label=\\\"kgpd_O01\\\" fontsize=10];\\n  kgpd_Add [shape=box style=\\\"filled,rounded\\\" color=orange label=\\\"Add\\n(kgpd_Add)\\\" fontsize=10];\\n  kgpd_O02 -> kgpd_Add;\\n  kgpd_Addcst -> kgpd_Add;\\n  kgpd_Add -> kgpd_O01;\\n\\n  kgpd_O0 [shape=box label=\\\"kgpd_O0\\\" fontsize=10];\\n  kgpd_Identity [shape=box style=\\\"filled,rounded\\\" color=orange label=\\\"Identity\\n(kgpd_Identity)\\\" fontsize=10];\\n  kgpd_O01 -> kgpd_Identity;\\n  kgpd_Identity -> kgpd_O0;\\n\\n  gpr_O0 [shape=box label=\\\"gpr_O0\\\" fontsize=10];\\n  gpr_MatMul [shape=box style=\\\"filled,rounded\\\" color=orange label=\\\"MatMul\\n(gpr_MatMul)\\\" fontsize=10];\\n  kgpd_O0 -> gpr_MatMul;\\n  gpr_MatMulcst -> gpr_MatMul;\\n  gpr_MatMul -> gpr_O0;\\n\\n  gpr_Add [shape=box style=\\\"filled,rounded\\\" color=orange label=\\\"Add\\n(gpr_Add)\\\" fontsize=10];\\n  gpr_O0 -> gpr_Add;\\n  gpr_Addcst -> gpr_Add;\\n  gpr_Add -> GPmean;\\n}\");\n",
-              "document.getElementById('M1c20a2ae95eb48c58d8ebc34dc0793da').innerHTML = svgGraph; });\n",
+              "require(['http://www.xavierdupre.fr/js/vizjs/viz.js'], function() { var svgGraph = Viz(\"digraph{\\n  ranksep=0.25;\\n  orientation=portrait;\\n  nodesep=0.05;\\n\\n  X [shape=box color=red label=\\\"X\\ndouble((0, 13))\\\" fontsize=10];\\n\\n  GPmean [shape=box color=green label=\\\"GPmean\\ndouble((0, 1))\\\" fontsize=10];\\n\\n  kgpd_MatMulcst [shape=box label=\\\"kgpd_MatMulcst\\nfloat64((13, 379))\\n[[2.98190e-01 1.10874e+01 8.22100e-02 ... 8.82900e...\\\" fontsize=10];\\n  kgpd_Addcst [shape=box label=\\\"kgpd_Addcst\\nfloat64((1,))\\n[705.79396344]\\\" fontsize=10];\\n  gpr_MatMulcst [shape=box label=\\\"gpr_MatMulcst\\nfloat64((379,))\\n[ 1.06765100e+00 -3.89747603e-01  4.00184291e-01 -...\\\" fontsize=10];\\n  gpr_Addcst [shape=box label=\\\"gpr_Addcst\\nfloat64((1, 1))\\n[[0.]]\\\" fontsize=10];\\n\\n  kgpd_Y0 [shape=box label=\\\"kgpd_Y0\\\" fontsize=10];\\n  kgpd_MatMul [shape=box style=\\\"filled,rounded\\\" color=orange label=\\\"MatMul\\n(kgpd_MatMul)\\\" fontsize=10];\\n  X -> kgpd_MatMul;\\n  kgpd_MatMulcst -> kgpd_MatMul;\\n  kgpd_MatMul -> kgpd_Y0;\\n\\n  kgpd_C0 [shape=box label=\\\"kgpd_C0\\\" fontsize=10];\\n  kgpd_Add [shape=box style=\\\"filled,rounded\\\" color=orange label=\\\"Add\\n(kgpd_Add)\\\" fontsize=10];\\n  kgpd_Y0 -> kgpd_Add;\\n  kgpd_Addcst -> kgpd_Add;\\n  kgpd_Add -> kgpd_C0;\\n\\n  kgpd_output0 [shape=box label=\\\"kgpd_output0\\\" fontsize=10];\\n  kgpd_Identity [shape=box style=\\\"filled,rounded\\\" color=orange label=\\\"Identity\\n(kgpd_Identity)\\\" fontsize=10];\\n  kgpd_C0 -> kgpd_Identity;\\n  kgpd_Identity -> kgpd_output0;\\n\\n  gpr_Y0 [shape=box label=\\\"gpr_Y0\\\" fontsize=10];\\n  gpr_MatMul [shape=box style=\\\"filled,rounded\\\" color=orange label=\\\"MatMul\\n(gpr_MatMul)\\\" fontsize=10];\\n  kgpd_output0 -> gpr_MatMul;\\n  gpr_MatMulcst -> gpr_MatMul;\\n  gpr_MatMul -> gpr_Y0;\\n\\n  gpr_Add [shape=box style=\\\"filled,rounded\\\" color=orange label=\\\"Add\\n(gpr_Add)\\\" fontsize=10];\\n  gpr_Y0 -> gpr_Add;\\n  gpr_Addcst -> gpr_Add;\\n  gpr_Add -> GPmean;\\n}\");\n",
+              "document.getElementById('Md6d6f79f149e4869a5c902cd901836e4').innerHTML = svgGraph; });\n",
               "\n",
               "</script>"
             ],
             "text/plain": [
-              "<jyquickhelper.jspy.render_nb_js_dot.RenderJsDot at 0x2d1f41c25f8>"
+              "<jyquickhelper.jspy.render_nb_js_dot.RenderJsDot at 0x2c04cd1a828>"
             ]
           },
           "execution_count": 24,
@@ -822,205 +815,205 @@
               "  <thead>\n",
               "    <tr style=\"text-align: right;\">\n",
               "      <th></th>\n",
-              "      <th>cmp</th>\n",
               "      <th>metric</th>\n",
-              "      <th>name</th>\n",
-              "      <th>shape[0]</th>\n",
-              "      <th>shape[1]</th>\n",
               "      <th>step</th>\n",
               "      <th>v[0]</th>\n",
               "      <th>v[1]</th>\n",
+              "      <th>cmp</th>\n",
+              "      <th>name</th>\n",
               "      <th>value[0]</th>\n",
+              "      <th>shape[0]</th>\n",
               "      <th>value[1]</th>\n",
+              "      <th>shape[1]</th>\n",
               "    </tr>\n",
               "  </thead>\n",
               "  <tbody>\n",
               "    <tr>\n",
               "      <th>0</th>\n",
-              "      <td>OK</td>\n",
               "      <td>nb_results</td>\n",
-              "      <td>NaN</td>\n",
-              "      <td>NaN</td>\n",
-              "      <td>NaN</td>\n",
               "      <td>-1</td>\n",
               "      <td>10</td>\n",
               "      <td>1.000000e+01</td>\n",
+              "      <td>OK</td>\n",
+              "      <td>NaN</td>\n",
+              "      <td>NaN</td>\n",
+              "      <td>NaN</td>\n",
               "      <td>NaN</td>\n",
               "      <td>NaN</td>\n",
               "    </tr>\n",
               "    <tr>\n",
               "      <th>1</th>\n",
-              "      <td>OK</td>\n",
               "      <td>abs-diff</td>\n",
+              "      <td>0</td>\n",
+              "      <td>0</td>\n",
+              "      <td>4.893252e-08</td>\n",
+              "      <td>OK</td>\n",
               "      <td>X</td>\n",
+              "      <td>[[0.06263, 0.0, 11.93, 0.0, 0.573, 6.593, 69.1...</td>\n",
               "      <td>(127, 13)</td>\n",
+              "      <td>[[0.06263, 0.0, 11.93, 0.0, 0.573, 6.593, 69.1...</td>\n",
               "      <td>(127, 13)</td>\n",
-              "      <td>0</td>\n",
-              "      <td>0</td>\n",
-              "      <td>5.553783e-08</td>\n",
-              "      <td>[[0.03768, 80.0, 1.52, 0.0, 0.404, 7.274, 38.3...</td>\n",
-              "      <td>[[0.03768, 80.0, 1.52, 0.0, 0.404, 7.274, 38.3...</td>\n",
               "    </tr>\n",
               "    <tr>\n",
               "      <th>2</th>\n",
-              "      <td>OK</td>\n",
               "      <td>abs-diff</td>\n",
-              "      <td>kgpd_MatMulcst</td>\n",
-              "      <td>(13, 379)</td>\n",
-              "      <td>(13, 379)</td>\n",
               "      <td>1</td>\n",
               "      <td>0</td>\n",
-              "      <td>5.647917e-08</td>\n",
-              "      <td>[[0.1712, 0.07896, 0.08308, 0.04011, 0.21977, ...</td>\n",
-              "      <td>[[0.1712, 0.07896, 0.08308, 0.04011, 0.21977, ...</td>\n",
+              "      <td>2.615161e-02</td>\n",
+              "      <td>e&lt;0.1</td>\n",
+              "      <td>GPmean</td>\n",
+              "      <td>[[24.375, 20.625, 28.5, 18.3125, 18.15625, 15....</td>\n",
+              "      <td>(1, 127)</td>\n",
+              "      <td>[[24.682750829830184, 20.17216288333293, 28.42...</td>\n",
+              "      <td>(1, 127)</td>\n",
               "    </tr>\n",
               "    <tr>\n",
               "      <th>3</th>\n",
-              "      <td>OK</td>\n",
               "      <td>abs-diff</td>\n",
-              "      <td>kgpd_Addcst</td>\n",
-              "      <td>(1,)</td>\n",
-              "      <td>(1,)</td>\n",
               "      <td>2</td>\n",
               "      <td>0</td>\n",
-              "      <td>4.773335e-08</td>\n",
-              "      <td>[1073.5477]</td>\n",
-              "      <td>[1073.5477807362217]</td>\n",
+              "      <td>5.553783e-08</td>\n",
+              "      <td>OK</td>\n",
+              "      <td>kgpd_MatMulcst</td>\n",
+              "      <td>[[0.29819, 11.0874, 0.08221, 0.12802, 0.62356,...</td>\n",
+              "      <td>(13, 379)</td>\n",
+              "      <td>[[0.29819, 11.0874, 0.08221, 0.12802, 0.62356,...</td>\n",
+              "      <td>(13, 379)</td>\n",
               "    </tr>\n",
               "    <tr>\n",
               "      <th>4</th>\n",
-              "      <td>OK</td>\n",
               "      <td>abs-diff</td>\n",
-              "      <td>gpr_MatMulcst</td>\n",
-              "      <td>(379,)</td>\n",
-              "      <td>(379,)</td>\n",
               "      <td>3</td>\n",
               "      <td>0</td>\n",
-              "      <td>5.481368e-08</td>\n",
-              "      <td>[0.2759209, -0.01826272, 0.30508244, -0.164773...</td>\n",
-              "      <td>[0.2759209068378013, -0.018262720128066828, 0....</td>\n",
+              "      <td>2.569000e-08</td>\n",
+              "      <td>OK</td>\n",
+              "      <td>kgpd_Addcst</td>\n",
+              "      <td>[705.79395]</td>\n",
+              "      <td>(1,)</td>\n",
+              "      <td>[705.7939634443445]</td>\n",
+              "      <td>(1,)</td>\n",
               "    </tr>\n",
               "    <tr>\n",
               "      <th>5</th>\n",
-              "      <td>OK</td>\n",
               "      <td>abs-diff</td>\n",
-              "      <td>gpr_Addcst</td>\n",
-              "      <td>(1, 1)</td>\n",
-              "      <td>(1, 1)</td>\n",
               "      <td>4</td>\n",
               "      <td>0</td>\n",
-              "      <td>0.000000e+00</td>\n",
-              "      <td>[[0.0]]</td>\n",
-              "      <td>[[0.0]]</td>\n",
+              "      <td>5.295014e-08</td>\n",
+              "      <td>OK</td>\n",
+              "      <td>gpr_MatMulcst</td>\n",
+              "      <td>[1.067651, -0.3897476, 0.4001843, -0.28861365,...</td>\n",
+              "      <td>(379,)</td>\n",
+              "      <td>[1.0676510043582548, -0.3897476028987512, 0.40...</td>\n",
+              "      <td>(379,)</td>\n",
               "    </tr>\n",
               "    <tr>\n",
               "      <th>6</th>\n",
-              "      <td>OK</td>\n",
               "      <td>abs-diff</td>\n",
-              "      <td>kgpd_O02</td>\n",
-              "      <td>(127, 379)</td>\n",
-              "      <td>(127, 379)</td>\n",
               "      <td>5</td>\n",
               "      <td>0</td>\n",
-              "      <td>1.729052e-07</td>\n",
-              "      <td>[[285506.28, 286446.3, 220675.92, 271909.88, 2...</td>\n",
-              "      <td>[[285506.252193816, 286446.3127387128, 220675....</td>\n",
+              "      <td>0.000000e+00</td>\n",
+              "      <td>OK</td>\n",
+              "      <td>gpr_Addcst</td>\n",
+              "      <td>[[0.0]]</td>\n",
+              "      <td>(1, 1)</td>\n",
+              "      <td>[[0.0]]</td>\n",
+              "      <td>(1, 1)</td>\n",
               "    </tr>\n",
               "    <tr>\n",
               "      <th>7</th>\n",
-              "      <td>OK</td>\n",
               "      <td>abs-diff</td>\n",
-              "      <td>kgpd_O01</td>\n",
-              "      <td>(127, 379)</td>\n",
-              "      <td>(127, 379)</td>\n",
               "      <td>6</td>\n",
               "      <td>0</td>\n",
-              "      <td>2.197289e-07</td>\n",
-              "      <td>[[286579.84, 287519.88, 221749.47, 272983.44, ...</td>\n",
-              "      <td>[[286579.79997455224, 287519.86051944905, 2217...</td>\n",
+              "      <td>2.003658e-07</td>\n",
+              "      <td>OK</td>\n",
+              "      <td>kgpd_Y0</td>\n",
+              "      <td>[[234216.27, 315237.94, 243189.44, 267891.22, ...</td>\n",
+              "      <td>(127, 379)</td>\n",
+              "      <td>[[234216.25805194405, 315237.96387384634, 2431...</td>\n",
+              "      <td>(127, 379)</td>\n",
               "    </tr>\n",
               "    <tr>\n",
               "      <th>8</th>\n",
-              "      <td>OK</td>\n",
               "      <td>abs-diff</td>\n",
-              "      <td>kgpd_O0</td>\n",
-              "      <td>(127, 379)</td>\n",
-              "      <td>(127, 379)</td>\n",
               "      <td>7</td>\n",
               "      <td>0</td>\n",
-              "      <td>2.197289e-07</td>\n",
-              "      <td>[[286579.84, 287519.88, 221749.47, 272983.44, ...</td>\n",
-              "      <td>[[286579.79997455224, 287519.86051944905, 2217...</td>\n",
+              "      <td>2.003658e-07</td>\n",
+              "      <td>OK</td>\n",
+              "      <td>kgpd_C0</td>\n",
+              "      <td>[[234216.27, 315237.94, 243189.44, 267891.22, ...</td>\n",
+              "      <td>(127, 379)</td>\n",
+              "      <td>[[234216.25805194405, 315237.96387384634, 2431...</td>\n",
+              "      <td>(127, 379)</td>\n",
               "    </tr>\n",
               "    <tr>\n",
               "      <th>9</th>\n",
-              "      <td>e&lt;0.1</td>\n",
               "      <td>abs-diff</td>\n",
-              "      <td>gpr_O0</td>\n",
-              "      <td>(127,)</td>\n",
-              "      <td>(127,)</td>\n",
               "      <td>8</td>\n",
               "      <td>0</td>\n",
-              "      <td>6.591271e-02</td>\n",
-              "      <td>[33.125, 6.625, 23.5, 22.625, 10.25, 19.75, 13...</td>\n",
-              "      <td>[33.162191349547356, 6.8058363744057715, 23.80...</td>\n",
+              "      <td>2.003658e-07</td>\n",
+              "      <td>OK</td>\n",
+              "      <td>kgpd_output0</td>\n",
+              "      <td>[[234216.27, 315237.94, 243189.44, 267891.22, ...</td>\n",
+              "      <td>(127, 379)</td>\n",
+              "      <td>[[234216.25805194405, 315237.96387384634, 2431...</td>\n",
+              "      <td>(127, 379)</td>\n",
               "    </tr>\n",
               "    <tr>\n",
               "      <th>10</th>\n",
-              "      <td>e&lt;0.1</td>\n",
               "      <td>abs-diff</td>\n",
-              "      <td>GPmean</td>\n",
-              "      <td>(1, 127)</td>\n",
-              "      <td>(1, 127)</td>\n",
               "      <td>9</td>\n",
               "      <td>0</td>\n",
-              "      <td>6.591271e-02</td>\n",
-              "      <td>[[33.125, 6.625, 23.5, 22.625, 10.25, 19.75, 1...</td>\n",
-              "      <td>[[33.162191349547356, 6.8058363744057715, 23.8...</td>\n",
+              "      <td>2.615161e-02</td>\n",
+              "      <td>e&lt;0.1</td>\n",
+              "      <td>gpr_Y0</td>\n",
+              "      <td>[24.375, 20.625, 28.5, 18.3125, 18.15625, 15.5...</td>\n",
+              "      <td>(127,)</td>\n",
+              "      <td>[24.682750829830184, 20.17216288333293, 28.428...</td>\n",
+              "      <td>(127,)</td>\n",
               "    </tr>\n",
               "  </tbody>\n",
               "</table>\n",
               "</div>"
             ],
             "text/plain": [
-              "      cmp      metric            name    shape[0]    shape[1]  step  v[0]  \\\n",
-              "0      OK  nb_results             NaN         NaN         NaN    -1    10   \n",
-              "1      OK    abs-diff               X   (127, 13)   (127, 13)     0     0   \n",
-              "2      OK    abs-diff  kgpd_MatMulcst   (13, 379)   (13, 379)     1     0   \n",
-              "3      OK    abs-diff     kgpd_Addcst        (1,)        (1,)     2     0   \n",
-              "4      OK    abs-diff   gpr_MatMulcst      (379,)      (379,)     3     0   \n",
-              "5      OK    abs-diff      gpr_Addcst      (1, 1)      (1, 1)     4     0   \n",
-              "6      OK    abs-diff        kgpd_O02  (127, 379)  (127, 379)     5     0   \n",
-              "7      OK    abs-diff        kgpd_O01  (127, 379)  (127, 379)     6     0   \n",
-              "8      OK    abs-diff         kgpd_O0  (127, 379)  (127, 379)     7     0   \n",
-              "9   e<0.1    abs-diff          gpr_O0      (127,)      (127,)     8     0   \n",
-              "10  e<0.1    abs-diff          GPmean    (1, 127)    (1, 127)     9     0   \n",
+              "        metric  step  v[0]          v[1]    cmp            name  \\\n",
+              "0   nb_results    -1    10  1.000000e+01     OK             NaN   \n",
+              "1     abs-diff     0     0  4.893252e-08     OK               X   \n",
+              "2     abs-diff     1     0  2.615161e-02  e<0.1          GPmean   \n",
+              "3     abs-diff     2     0  5.553783e-08     OK  kgpd_MatMulcst   \n",
+              "4     abs-diff     3     0  2.569000e-08     OK     kgpd_Addcst   \n",
+              "5     abs-diff     4     0  5.295014e-08     OK   gpr_MatMulcst   \n",
+              "6     abs-diff     5     0  0.000000e+00     OK      gpr_Addcst   \n",
+              "7     abs-diff     6     0  2.003658e-07     OK         kgpd_Y0   \n",
+              "8     abs-diff     7     0  2.003658e-07     OK         kgpd_C0   \n",
+              "9     abs-diff     8     0  2.003658e-07     OK    kgpd_output0   \n",
+              "10    abs-diff     9     0  2.615161e-02  e<0.1          gpr_Y0   \n",
               "\n",
-              "            v[1]                                           value[0]  \\\n",
-              "0   1.000000e+01                                                NaN   \n",
-              "1   5.553783e-08  [[0.03768, 80.0, 1.52, 0.0, 0.404, 7.274, 38.3...   \n",
-              "2   5.647917e-08  [[0.1712, 0.07896, 0.08308, 0.04011, 0.21977, ...   \n",
-              "3   4.773335e-08                                        [1073.5477]   \n",
-              "4   5.481368e-08  [0.2759209, -0.01826272, 0.30508244, -0.164773...   \n",
-              "5   0.000000e+00                                            [[0.0]]   \n",
-              "6   1.729052e-07  [[285506.28, 286446.3, 220675.92, 271909.88, 2...   \n",
-              "7   2.197289e-07  [[286579.84, 287519.88, 221749.47, 272983.44, ...   \n",
-              "8   2.197289e-07  [[286579.84, 287519.88, 221749.47, 272983.44, ...   \n",
-              "9   6.591271e-02  [33.125, 6.625, 23.5, 22.625, 10.25, 19.75, 13...   \n",
-              "10  6.591271e-02  [[33.125, 6.625, 23.5, 22.625, 10.25, 19.75, 1...   \n",
+              "                                             value[0]    shape[0]  \\\n",
+              "0                                                 NaN         NaN   \n",
+              "1   [[0.06263, 0.0, 11.93, 0.0, 0.573, 6.593, 69.1...   (127, 13)   \n",
+              "2   [[24.375, 20.625, 28.5, 18.3125, 18.15625, 15....    (1, 127)   \n",
+              "3   [[0.29819, 11.0874, 0.08221, 0.12802, 0.62356,...   (13, 379)   \n",
+              "4                                         [705.79395]        (1,)   \n",
+              "5   [1.067651, -0.3897476, 0.4001843, -0.28861365,...      (379,)   \n",
+              "6                                             [[0.0]]      (1, 1)   \n",
+              "7   [[234216.27, 315237.94, 243189.44, 267891.22, ...  (127, 379)   \n",
+              "8   [[234216.27, 315237.94, 243189.44, 267891.22, ...  (127, 379)   \n",
+              "9   [[234216.27, 315237.94, 243189.44, 267891.22, ...  (127, 379)   \n",
+              "10  [24.375, 20.625, 28.5, 18.3125, 18.15625, 15.5...      (127,)   \n",
               "\n",
-              "                                             value[1]  \n",
-              "0                                                 NaN  \n",
-              "1   [[0.03768, 80.0, 1.52, 0.0, 0.404, 7.274, 38.3...  \n",
-              "2   [[0.1712, 0.07896, 0.08308, 0.04011, 0.21977, ...  \n",
-              "3                                [1073.5477807362217]  \n",
-              "4   [0.2759209068378013, -0.018262720128066828, 0....  \n",
-              "5                                             [[0.0]]  \n",
-              "6   [[285506.252193816, 286446.3127387128, 220675....  \n",
-              "7   [[286579.79997455224, 287519.86051944905, 2217...  \n",
-              "8   [[286579.79997455224, 287519.86051944905, 2217...  \n",
-              "9   [33.162191349547356, 6.8058363744057715, 23.80...  \n",
-              "10  [[33.162191349547356, 6.8058363744057715, 23.8...  "
+              "                                             value[1]    shape[1]  \n",
+              "0                                                 NaN         NaN  \n",
+              "1   [[0.06263, 0.0, 11.93, 0.0, 0.573, 6.593, 69.1...   (127, 13)  \n",
+              "2   [[24.682750829830184, 20.17216288333293, 28.42...    (1, 127)  \n",
+              "3   [[0.29819, 11.0874, 0.08221, 0.12802, 0.62356,...   (13, 379)  \n",
+              "4                                 [705.7939634443445]        (1,)  \n",
+              "5   [1.0676510043582548, -0.3897476028987512, 0.40...      (379,)  \n",
+              "6                                             [[0.0]]      (1, 1)  \n",
+              "7   [[234216.25805194405, 315237.96387384634, 2431...  (127, 379)  \n",
+              "8   [[234216.25805194405, 315237.96387384634, 2431...  (127, 379)  \n",
+              "9   [[234216.25805194405, 315237.96387384634, 2431...  (127, 379)  \n",
+              "10  [24.682750829830184, 20.17216288333293, 28.428...      (127,)  "
             ]
           },
           "execution_count": 27,
@@ -1053,7 +1046,7 @@
       "outputs": [
         {
           "data": {
-            "image/png": "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\n",
+            "image/png": "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\n",
             "text/plain": [
               "<Figure size 1008x288 with 1 Axes>"
             ]
@@ -1085,8 +1078,8 @@
         {
           "data": {
             "text/plain": [
-              "array([1.22073106e-06, 1.26985833e-06, 1.29081309e-06, 1.53481960e-06,\n",
-              "       1.54972076e-06])"
+              "array([1.51888526e-06, 1.52538996e-06, 1.53196743e-06, 1.70313433e-06,\n",
+              "       1.89113780e-06])"
             ]
           },
           "execution_count": 29,
@@ -1115,8 +1108,8 @@
         {
           "data": {
             "text/plain": [
-              "array([1.22073106e-06, 1.26985833e-06, 1.29081309e-06, 1.53481960e-06,\n",
-              "       1.54972076e-06])"
+              "array([1.51888526e-06, 1.52538996e-06, 1.53196743e-06, 1.70313433e-06,\n",
+              "       1.89113780e-06])"
             ]
           },
           "execution_count": 30,
@@ -1144,7 +1137,7 @@
       "outputs": [
         {
           "data": {
-            "image/png": "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\n",
+            "image/png": "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\n",
             "text/plain": [
               "<Figure size 1008x288 with 1 Axes>"
             ]
@@ -1164,358 +1157,10 @@
         "             \"\\nfor a GaussianProcessRegressor\");"
       ]
     },
-    {
-      "cell_type": "markdown",
-      "metadata": {},
-      "source": [
-        "## Partial use of float64"
-      ]
-    },
     {
       "cell_type": "code",
       "execution_count": 31,
       "metadata": {},
-      "outputs": [
-        {
-          "data": {
-            "text/html": [
-              "<div id=\"M4759a741de104ea1ac07ac636ca00e8a-cont\"><div id=\"M4759a741de104ea1ac07ac636ca00e8a\" style=\"width:100%;height:100%;\"></div></div>\n",
-              "<script>\n",
-              "\n",
-              "require(['http://www.xavierdupre.fr/js/vizjs/viz.js'], function() { var svgGraph = Viz(\"digraph{\\n  nodesep=0.05;\\n  ranksep=0.25;\\n  orientation=portrait;\\n\\n  X [shape=box color=red label=\\\"X\\nfloat((1, 13))\\\" fontsize=10];\\n\\n  GPmean [shape=box color=green label=\\\"GPmean\\nfloat((1, 1))\\\" fontsize=10];\\n\\n  kgpd_MatMulcst [shape=box label=\\\"kgpd_MatMulcst\\nfloat64((13, 379))\\n[[1.71200007e-01 7.89600015e-0...\\\" fontsize=10];\\n  kgpd_Addcst [shape=box label=\\\"kgpd_Addcst\\nfloat32((1,))\\n[1073.5477]\\\" fontsize=10];\\n  gpr_MatMulcst [shape=box label=\\\"gpr_MatMulcst\\nfloat64((379,))\\n[ 2.75920907e-01 -1.82627201e-...\\\" fontsize=10];\\n  gpr_Addcst [shape=box label=\\\"gpr_Addcst\\nfloat32((1, 1))\\n[[0.]]\\\" fontsize=10];\\n\\n  kgpd_O04 [shape=box label=\\\"kgpd_O04\\\" fontsize=10];\\n  kgpd_Cast [shape=box style=\\\"filled,rounded\\\" color=orange label=\\\"Cast\\n(kgpd_Cast)\\nto=11\\\" fontsize=10];\\n  X -> kgpd_Cast;\\n  kgpd_Cast -> kgpd_O04;\\n\\n  kgpd_O03 [shape=box label=\\\"kgpd_O03\\\" fontsize=10];\\n  kgpd_MatMul [shape=box style=\\\"filled,rounded\\\" color=orange label=\\\"MatMul\\n(kgpd_MatMul)\\\" fontsize=10];\\n  kgpd_O04 -> kgpd_MatMul;\\n  kgpd_MatMulcst -> kgpd_MatMul;\\n  kgpd_MatMul -> kgpd_O03;\\n\\n  kgpd_O02 [shape=box label=\\\"kgpd_O02\\\" fontsize=10];\\n  kgpd_Cast1 [shape=box style=\\\"filled,rounded\\\" color=orange label=\\\"Cast\\n(kgpd_Cast1)\\nto=1\\\" fontsize=10];\\n  kgpd_O03 -> kgpd_Cast1;\\n  kgpd_Cast1 -> kgpd_O02;\\n\\n  kgpd_O01 [shape=box label=\\\"kgpd_O01\\\" fontsize=10];\\n  kgpd_Add [shape=box style=\\\"filled,rounded\\\" color=orange label=\\\"Add\\n(kgpd_Add)\\\" fontsize=10];\\n  kgpd_O02 -> kgpd_Add;\\n  kgpd_Addcst -> kgpd_Add;\\n  kgpd_Add -> kgpd_O01;\\n\\n  kgpd_O0 [shape=box label=\\\"kgpd_O0\\\" fontsize=10];\\n  kgpd_Identity [shape=box style=\\\"filled,rounded\\\" color=orange label=\\\"Identity\\n(kgpd_Identity)\\\" fontsize=10];\\n  kgpd_O01 -> kgpd_Identity;\\n  kgpd_Identity -> kgpd_O0;\\n\\n  gpr_O02 [shape=box label=\\\"gpr_O02\\\" fontsize=10];\\n  gpr_Cast [shape=box style=\\\"filled,rounded\\\" color=orange label=\\\"Cast\\n(gpr_Cast)\\nto=11\\\" fontsize=10];\\n  kgpd_O0 -> gpr_Cast;\\n  gpr_Cast -> gpr_O02;\\n\\n  gpr_O01 [shape=box label=\\\"gpr_O01\\\" fontsize=10];\\n  gpr_MatMul [shape=box style=\\\"filled,rounded\\\" color=orange label=\\\"MatMul\\n(gpr_MatMul)\\\" fontsize=10];\\n  gpr_O02 -> gpr_MatMul;\\n  gpr_MatMulcst -> gpr_MatMul;\\n  gpr_MatMul -> gpr_O01;\\n\\n  gpr_O0 [shape=box label=\\\"gpr_O0\\\" fontsize=10];\\n  gpr_Cast1 [shape=box style=\\\"filled,rounded\\\" color=orange label=\\\"Cast\\n(gpr_Cast1)\\nto=1\\\" fontsize=10];\\n  gpr_O01 -> gpr_Cast1;\\n  gpr_Cast1 -> gpr_O0;\\n\\n  gpr_Add [shape=box style=\\\"filled,rounded\\\" color=orange label=\\\"Add\\n(gpr_Add)\\\" fontsize=10];\\n  gpr_O0 -> gpr_Add;\\n  gpr_Addcst -> gpr_Add;\\n  gpr_Add -> GPmean;\\n}\");\n",
-              "document.getElementById('M4759a741de104ea1ac07ac636ca00e8a').innerHTML = svgGraph; });\n",
-              "\n",
-              "</script>"
-            ],
-            "text/plain": [
-              "<jyquickhelper.jspy.render_nb_js_dot.RenderJsDot at 0x2d1f4d96470>"
-            ]
-          },
-          "execution_count": 32,
-          "metadata": {},
-          "output_type": "execute_result"
-        }
-      ],
-      "source": [
-        "onnxgau48 = to_onnx(gau, X_train.astype(numpy.float32), dtype=numpy.float32,\n",
-        "                    options={GaussianProcessRegressor: {'float64': True}})\n",
-        "%onnxview onnxgau48"
-      ]
-    },
-    {
-      "cell_type": "code",
-      "execution_count": 32,
-      "metadata": {},
-      "outputs": [
-        {
-          "data": {
-            "text/plain": [
-              "array([0.16995945, 0.17210525, 0.20095541, 0.20111441, 0.24590116])"
-            ]
-          },
-          "execution_count": 33,
-          "metadata": {},
-          "output_type": "execute_result"
-        }
-      ],
-      "source": [
-        "oinf48 = OnnxInference(onnxgau48, runtime=\"python\")\n",
-        "ort_pred48 = oinf48.run({'X': X_test.astype(numpy.float32)})['GPmean']\n",
-        "numpy.sort(numpy.sort(numpy.squeeze(ort_pred48 - ort_pred64)))[-5:]"
-      ]
-    },
-    {
-      "cell_type": "code",
-      "execution_count": 33,
-      "metadata": {},
-      "outputs": [
-        {
-          "data": {
-            "image/png": "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\n",
-            "text/plain": [
-              "<Figure size 1008x288 with 1 Axes>"
-            ]
-          },
-          "metadata": {
-            "needs_background": "light"
-          },
-          "output_type": "display_data"
-        }
-      ],
-      "source": [
-        "sbs = side_by_side_by_values([(oinf48, {'X': X_test.astype(numpy.float32)}),\n",
-        "                              (oinf64, {'X': X_test.astype(numpy.float64)})])\n",
-        "df = DataFrame(sbs)\n",
-        "ax = df[['name', 'v[1]']].iloc[1:].set_index('name').plot(kind='bar', figsize=(14,4), logy=True)\n",
-        "ax.set_title(\"Relative differences for each output between float64 and float64 rounded to float32\"\n",
-        "             \"\\nfor a GaussianProcessRegressor\");"
-      ]
-    },
-    {
-      "cell_type": "code",
-      "execution_count": 34,
-      "metadata": {},
-      "outputs": [
-        {
-          "data": {
-            "text/html": [
-              "<div>\n",
-              "<style scoped>\n",
-              "    .dataframe tbody tr th:only-of-type {\n",
-              "        vertical-align: middle;\n",
-              "    }\n",
-              "\n",
-              "    .dataframe tbody tr th {\n",
-              "        vertical-align: top;\n",
-              "    }\n",
-              "\n",
-              "    .dataframe thead th {\n",
-              "        text-align: right;\n",
-              "    }\n",
-              "</style>\n",
-              "<table border=\"1\" class=\"dataframe\">\n",
-              "  <thead>\n",
-              "    <tr style=\"text-align: right;\">\n",
-              "      <th></th>\n",
-              "      <th>cmp</th>\n",
-              "      <th>metric</th>\n",
-              "      <th>name</th>\n",
-              "      <th>names</th>\n",
-              "      <th>shape[0]</th>\n",
-              "      <th>shape[1]</th>\n",
-              "      <th>step</th>\n",
-              "      <th>v[0]</th>\n",
-              "      <th>v[1]</th>\n",
-              "      <th>value[0]</th>\n",
-              "      <th>value[1]</th>\n",
-              "    </tr>\n",
-              "  </thead>\n",
-              "  <tbody>\n",
-              "    <tr>\n",
-              "      <th>0</th>\n",
-              "      <td>!=</td>\n",
-              "      <td>nb_results</td>\n",
-              "      <td>NaN</td>\n",
-              "      <td>NaN</td>\n",
-              "      <td>NaN</td>\n",
-              "      <td>NaN</td>\n",
-              "      <td>-1</td>\n",
-              "      <td>14</td>\n",
-              "      <td>1.000000e+01</td>\n",
-              "      <td>NaN</td>\n",
-              "      <td>NaN</td>\n",
-              "    </tr>\n",
-              "    <tr>\n",
-              "      <th>1</th>\n",
-              "      <td>OK</td>\n",
-              "      <td>abs-diff</td>\n",
-              "      <td>X</td>\n",
-              "      <td>NaN</td>\n",
-              "      <td>(127, 13)</td>\n",
-              "      <td>(127, 13)</td>\n",
-              "      <td>0</td>\n",
-              "      <td>0</td>\n",
-              "      <td>5.553783e-08</td>\n",
-              "      <td>[[0.03768, 80.0, 1.52, 0.0, 0.404, 7.274, 38.3...</td>\n",
-              "      <td>[[0.03768, 80.0, 1.52, 0.0, 0.404, 7.274, 38.3...</td>\n",
-              "    </tr>\n",
-              "    <tr>\n",
-              "      <th>2</th>\n",
-              "      <td>OK</td>\n",
-              "      <td>abs-diff</td>\n",
-              "      <td>kgpd_MatMulcst</td>\n",
-              "      <td>NaN</td>\n",
-              "      <td>(13, 379)</td>\n",
-              "      <td>(13, 379)</td>\n",
-              "      <td>1</td>\n",
-              "      <td>0</td>\n",
-              "      <td>5.647917e-08</td>\n",
-              "      <td>[[0.1712000072002411, 0.07896000146865845, 0.0...</td>\n",
-              "      <td>[[0.1712, 0.07896, 0.08308, 0.04011, 0.21977, ...</td>\n",
-              "    </tr>\n",
-              "    <tr>\n",
-              "      <th>3</th>\n",
-              "      <td>OK</td>\n",
-              "      <td>abs-diff</td>\n",
-              "      <td>kgpd_Addcst</td>\n",
-              "      <td>NaN</td>\n",
-              "      <td>(1,)</td>\n",
-              "      <td>(1,)</td>\n",
-              "      <td>2</td>\n",
-              "      <td>0</td>\n",
-              "      <td>4.773335e-08</td>\n",
-              "      <td>[1073.5477]</td>\n",
-              "      <td>[1073.5477807362217]</td>\n",
-              "    </tr>\n",
-              "    <tr>\n",
-              "      <th>4</th>\n",
-              "      <td>OK</td>\n",
-              "      <td>abs-diff</td>\n",
-              "      <td>gpr_MatMulcst</td>\n",
-              "      <td>NaN</td>\n",
-              "      <td>(379,)</td>\n",
-              "      <td>(379,)</td>\n",
-              "      <td>3</td>\n",
-              "      <td>0</td>\n",
-              "      <td>0.000000e+00</td>\n",
-              "      <td>[0.2759209068378013, -0.018262720128066828, 0....</td>\n",
-              "      <td>[0.2759209068378013, -0.018262720128066828, 0....</td>\n",
-              "    </tr>\n",
-              "    <tr>\n",
-              "      <th>5</th>\n",
-              "      <td>OK</td>\n",
-              "      <td>abs-diff</td>\n",
-              "      <td>gpr_Addcst</td>\n",
-              "      <td>NaN</td>\n",
-              "      <td>(1, 1)</td>\n",
-              "      <td>(1, 1)</td>\n",
-              "      <td>4</td>\n",
-              "      <td>0</td>\n",
-              "      <td>0.000000e+00</td>\n",
-              "      <td>[[0.0]]</td>\n",
-              "      <td>[[0.0]]</td>\n",
-              "    </tr>\n",
-              "    <tr>\n",
-              "      <th>6</th>\n",
-              "      <td>ERROR-&gt;=100000000000.0</td>\n",
-              "      <td>abs-diff</td>\n",
-              "      <td>NaN</td>\n",
-              "      <td>kgpd_O02 -&gt; kgpd_O04</td>\n",
-              "      <td>(127, 13)</td>\n",
-              "      <td>(127, 379)</td>\n",
-              "      <td>5</td>\n",
-              "      <td>0</td>\n",
-              "      <td>1.000000e+11</td>\n",
-              "      <td>[[0.0376800000667572, 80.0, 1.5199999809265137...</td>\n",
-              "      <td>[[285506.252193816, 286446.3127387128, 220675....</td>\n",
-              "    </tr>\n",
-              "    <tr>\n",
-              "      <th>7</th>\n",
-              "      <td>e&lt;0.01</td>\n",
-              "      <td>abs-diff</td>\n",
-              "      <td>NaN</td>\n",
-              "      <td>kgpd_O01 -&gt; kgpd_O03</td>\n",
-              "      <td>(127, 379)</td>\n",
-              "      <td>(127, 379)</td>\n",
-              "      <td>6</td>\n",
-              "      <td>0</td>\n",
-              "      <td>3.969042e-03</td>\n",
-              "      <td>[[285506.26227946434, 286446.3228386466, 22067...</td>\n",
-              "      <td>[[286579.79997455224, 287519.86051944905, 2217...</td>\n",
-              "    </tr>\n",
-              "    <tr>\n",
-              "      <th>8</th>\n",
-              "      <td>e&lt;0.01</td>\n",
-              "      <td>abs-diff</td>\n",
-              "      <td>NaN</td>\n",
-              "      <td>kgpd_O0 -&gt; kgpd_O02</td>\n",
-              "      <td>(127, 379)</td>\n",
-              "      <td>(127, 379)</td>\n",
-              "      <td>7</td>\n",
-              "      <td>0</td>\n",
-              "      <td>3.969088e-03</td>\n",
-              "      <td>[[285506.25, 286446.3, 220675.9, 271909.88, 23...</td>\n",
-              "      <td>[[286579.79997455224, 287519.86051944905, 2217...</td>\n",
-              "    </tr>\n",
-              "    <tr>\n",
-              "      <th>9</th>\n",
-              "      <td>ERROR-&gt;=100000000000.0</td>\n",
-              "      <td>abs-diff</td>\n",
-              "      <td>NaN</td>\n",
-              "      <td>gpr_O0 -&gt; kgpd_O01</td>\n",
-              "      <td>(127, 379)</td>\n",
-              "      <td>(127,)</td>\n",
-              "      <td>8</td>\n",
-              "      <td>0</td>\n",
-              "      <td>1.000000e+11</td>\n",
-              "      <td>[[286579.8, 287519.88, 221749.45, 272983.44, 2...</td>\n",
-              "      <td>[33.162191349547356, 6.8058363744057715, 23.80...</td>\n",
-              "    </tr>\n",
-              "    <tr>\n",
-              "      <th>10</th>\n",
-              "      <td>ERROR-&gt;=100000000000.0</td>\n",
-              "      <td>abs-diff</td>\n",
-              "      <td>NaN</td>\n",
-              "      <td>GPmean -&gt; kgpd_O0</td>\n",
-              "      <td>(127, 379)</td>\n",
-              "      <td>(1, 127)</td>\n",
-              "      <td>9</td>\n",
-              "      <td>0</td>\n",
-              "      <td>1.000000e+11</td>\n",
-              "      <td>[[286579.8, 287519.88, 221749.45, 272983.44, 2...</td>\n",
-              "      <td>[[33.162191349547356, 6.8058363744057715, 23.8...</td>\n",
-              "    </tr>\n",
-              "  </tbody>\n",
-              "</table>\n",
-              "</div>"
-            ],
-            "text/plain": [
-              "                       cmp      metric            name                 names  \\\n",
-              "0                       !=  nb_results             NaN                   NaN   \n",
-              "1                       OK    abs-diff               X                   NaN   \n",
-              "2                       OK    abs-diff  kgpd_MatMulcst                   NaN   \n",
-              "3                       OK    abs-diff     kgpd_Addcst                   NaN   \n",
-              "4                       OK    abs-diff   gpr_MatMulcst                   NaN   \n",
-              "5                       OK    abs-diff      gpr_Addcst                   NaN   \n",
-              "6   ERROR->=100000000000.0    abs-diff             NaN  kgpd_O02 -> kgpd_O04   \n",
-              "7                   e<0.01    abs-diff             NaN  kgpd_O01 -> kgpd_O03   \n",
-              "8                   e<0.01    abs-diff             NaN   kgpd_O0 -> kgpd_O02   \n",
-              "9   ERROR->=100000000000.0    abs-diff             NaN    gpr_O0 -> kgpd_O01   \n",
-              "10  ERROR->=100000000000.0    abs-diff             NaN     GPmean -> kgpd_O0   \n",
-              "\n",
-              "      shape[0]    shape[1]  step  v[0]          v[1]  \\\n",
-              "0          NaN         NaN    -1    14  1.000000e+01   \n",
-              "1    (127, 13)   (127, 13)     0     0  5.553783e-08   \n",
-              "2    (13, 379)   (13, 379)     1     0  5.647917e-08   \n",
-              "3         (1,)        (1,)     2     0  4.773335e-08   \n",
-              "4       (379,)      (379,)     3     0  0.000000e+00   \n",
-              "5       (1, 1)      (1, 1)     4     0  0.000000e+00   \n",
-              "6    (127, 13)  (127, 379)     5     0  1.000000e+11   \n",
-              "7   (127, 379)  (127, 379)     6     0  3.969042e-03   \n",
-              "8   (127, 379)  (127, 379)     7     0  3.969088e-03   \n",
-              "9   (127, 379)      (127,)     8     0  1.000000e+11   \n",
-              "10  (127, 379)    (1, 127)     9     0  1.000000e+11   \n",
-              "\n",
-              "                                             value[0]  \\\n",
-              "0                                                 NaN   \n",
-              "1   [[0.03768, 80.0, 1.52, 0.0, 0.404, 7.274, 38.3...   \n",
-              "2   [[0.1712000072002411, 0.07896000146865845, 0.0...   \n",
-              "3                                         [1073.5477]   \n",
-              "4   [0.2759209068378013, -0.018262720128066828, 0....   \n",
-              "5                                             [[0.0]]   \n",
-              "6   [[0.0376800000667572, 80.0, 1.5199999809265137...   \n",
-              "7   [[285506.26227946434, 286446.3228386466, 22067...   \n",
-              "8   [[285506.25, 286446.3, 220675.9, 271909.88, 23...   \n",
-              "9   [[286579.8, 287519.88, 221749.45, 272983.44, 2...   \n",
-              "10  [[286579.8, 287519.88, 221749.45, 272983.44, 2...   \n",
-              "\n",
-              "                                             value[1]  \n",
-              "0                                                 NaN  \n",
-              "1   [[0.03768, 80.0, 1.52, 0.0, 0.404, 7.274, 38.3...  \n",
-              "2   [[0.1712, 0.07896, 0.08308, 0.04011, 0.21977, ...  \n",
-              "3                                [1073.5477807362217]  \n",
-              "4   [0.2759209068378013, -0.018262720128066828, 0....  \n",
-              "5                                             [[0.0]]  \n",
-              "6   [[285506.252193816, 286446.3127387128, 220675....  \n",
-              "7   [[286579.79997455224, 287519.86051944905, 2217...  \n",
-              "8   [[286579.79997455224, 287519.86051944905, 2217...  \n",
-              "9   [33.162191349547356, 6.8058363744057715, 23.80...  \n",
-              "10  [[33.162191349547356, 6.8058363744057715, 23.8...  "
-            ]
-          },
-          "execution_count": 35,
-          "metadata": {},
-          "output_type": "execute_result"
-        }
-      ],
-      "source": [
-        "df"
-      ]
-    },
-    {
-      "cell_type": "code",
-      "execution_count": 35,
-      "metadata": {},
       "outputs": [],
       "source": []
     }
diff --git a/_unittests/ut_onnxrt/test_rt_valid_model_gaussian_process.py b/_unittests/ut_onnxrt/test_rt_valid_model_gaussian_process.py
index 0f60e28ca..6bf52e8fc 100644
--- a/_unittests/ut_onnxrt/test_rt_valid_model_gaussian_process.py
+++ b/_unittests/ut_onnxrt/test_rt_valid_model_gaussian_process.py
@@ -216,7 +216,8 @@ def test_partial_float64(self):
         X_train, X_test, y_train, _ = train_test_split(X, y)
         gau = GaussianProcessRegressor(alpha=10, kernel=DotProduct())
         gau.fit(X_train, y_train)
-        onnxgau48 = to_onnx(gau, X_train.astype(numpy.float32), dtype=numpy.float32)
+        onnxgau48 = to_onnx(gau, X_train.astype(
+            numpy.float32), dtype=numpy.float32)
         oinf48 = OnnxInference(onnxgau48, runtime="python")
         out = oinf48.run({'X': X_test.astype(numpy.float32)})
         y = out['GPmean']