diff --git a/notebooks/mushroom_analysis.ipynb b/notebooks/mushroom_analysis.ipynb index 8542f30..6149e3a 100644 --- a/notebooks/mushroom_analysis.ipynb +++ b/notebooks/mushroom_analysis.ipynb @@ -10,17 +10,282 @@ "name": "stdout", "output_type": "stream", "text": [ - "Notebook is ready\n" + "Feature data: (8124, 22)\n", + "Target data: (8124,)\n" + ] + }, + { + "data": { + "text/html": [ + "
\n", + "\n", + "\n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + "
cap_shapecap_surfacecap_colorbruisesodorgill_attachmentgill_spacinggill_sizegill_colorstalk_shape...stalk_surface_below_ringstalk_color_above_ringstalk_color_below_ringveil_typeveil_colorring_numberring_typespore-print-colorpopulationhabitat
0xsntpfcnke...swwpwopksu
1xsytafcbke...swwpwopnng
2bswtlfcbne...swwpwopnnm
3xywtpfcnne...swwpwopksu
4xsgfnfwbkt...swwpwoenag
\n", + "

5 rows × 22 columns

\n", + "
" + ], + "text/plain": [ + " cap_shape cap_surface cap_color bruises odor gill_attachment gill_spacing \\\n", + "0 x s n t p f c \n", + "1 x s y t a f c \n", + "2 b s w t l f c \n", + "3 x y w t p f c \n", + "4 x s g f n f w \n", + "\n", + " gill_size gill_color stalk_shape ... stalk_surface_below_ring \\\n", + "0 n k e ... s \n", + "1 b k e ... s \n", + "2 b n e ... s \n", + "3 n n e ... s \n", + "4 b k t ... s \n", + "\n", + " stalk_color_above_ring stalk_color_below_ring veil_type veil_color \\\n", + "0 w w p w \n", + "1 w w p w \n", + "2 w w p w \n", + "3 w w p w \n", + "4 w w p w \n", + "\n", + " ring_number ring_type spore-print-color population habitat \n", + "0 o p k s u \n", + "1 o p n n g \n", + "2 o p n n m \n", + "3 o p k s u \n", + "4 o e n a g \n", + "\n", + "[5 rows x 22 columns]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "text/plain": [ + "0 p\n", + "1 e\n", + "2 e\n", + "3 p\n", + "4 e\n", + "Name: class, dtype: category\n", + "Categories (2, str): ['e', 'p']" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "name": "stdout", + "output_type": "stream", + "text": [ + "class\n", + "e 4208\n", + "p 3916\n", + "Name: count, dtype: int64\n" ] } ], "source": [ - "import pandas as pd\n", - "import sklearn\n", - "import shap\n", - "import xgboost\n", + "from sklearn.datasets import fetch_openml\n", "\n", - "print(\"Notebook is ready\")" + "mushroom = fetch_openml(data_id=43922, as_frame=True)\n", + "\n", + "X = mushroom.data\n", + "y = mushroom.target\n", + "\n", + "print(\"Feature data:\", X.shape)\n", + "print(\"Target data:\", y.shape)\n", + "display(X.head())\n", + "display(y.head())\n", + "print(y.value_counts())" + ] + }, + { + "cell_type": "code", + "execution_count": 2, + "id": "69e1a254", + "metadata": {}, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "Missing cells: 0\n", + "Rows with missing cells: 0\n", + "stalk_root\n", + "b 3776\n", + "? 2480\n", + "e 1120\n", + "c 556\n", + "r 192\n", + "Name: count, dtype: int64\n" + ] + } + ], + "source": [ + "print(\"Missing cells:\", X.isna().sum().sum())\n", + "print(\"Rows with missing cells:\", X.isna().any(axis=1).sum())\n", + "print(X[\"stalk_root\"].value_counts(dropna=False))" ] } ],