sesi-10.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "71a76ded-ae03-49e5-ab58-4b082a9289c9",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "======= Binning =======\n",
      "Data Awal :\n",
      "     Number\n",
      "0       40\n",
      "1       34\n",
      "2       26\n",
      "3       32\n",
      "4       54\n",
      "5       54\n",
      "6       32\n",
      "7      523\n",
      "8       52\n",
      "9       13\n",
      "10      71\n",
      "11      33\n",
      "\n",
      "Data Setelah dilakukan Binning :\n",
      "     Number   bin\n",
      "0       40   Mid\n",
      "1       34   Mid\n",
      "2       26   Low\n",
      "3       32   Mid\n",
      "4       54   Mid\n",
      "5       54   Mid\n",
      "6       32   Mid\n",
      "7      523   NaN\n",
      "8       52   Mid\n",
      "9       13   Low\n",
      "10      71  High\n",
      "11      33   Mid\n",
      "\n",
      "\n",
      "======= Spliting =======\n",
      "Data Frame Awal\n",
      "       Name  Age\n",
      "0  Human A   40\n",
      "1  Human B   34\n",
      "2  Human C   26\n",
      "\n",
      "Data Frame setelah Split kolom 'Name', lalu menambahkan kedua kolom :\n",
      "       Name  Age  First Last\n",
      "0  Human A   40  Human    A\n",
      "1  Human B   34  Human    B\n",
      "2  Human C   26  Human    C\n"
     ]
    }
   ],
   "source": [
    "import pandas as pd\n",
    "\n",
    "'''\n",
    "Binning\n",
    "pengelompokan nilai seperti pemberian label pada range data yang di tentukan.\n",
    "Binning bisa membantu untuk menghindari overfitting. Namun binning juga\n",
    "mengorbankan informasi yang terkandung dari sebuah atribut sehingga,\n",
    "penggunaanya perlu dilakukan dengan teliti.\n",
    "'''\n",
    "\n",
    "print(\"======= Binning =======\")\n",
    "df = pd.DataFrame({'Number':[40, 34, 26, 32, 54, 54, 32, 523, 52, 13, 71, 33]})\n",
    "\n",
    "print(\"Data Awal :\\n\", df)\n",
    "df['bin'] = pd.cut(df['Number'], bins=[0, 30,  70, 100], labels=[\"Low\", \"Mid\", \"High\"])\n",
    "print(\"\\nData Setelah dilakukan Binning :\\n\", df)\n",
    "\n",
    "'''\n",
    "Spliting\n",
    "\n",
    "Memisahkan sebuah atribut menjadi atribut-atribut baru yang lebih mudah di\n",
    "pahami oleh model ML kita Contohnya seperti ini.\n",
    "Misal kita memiliki atribut yang berisi nama seseorang, katakanlah Satria Rahmat\n",
    "kita dapat membuat atribut baru yang berisi first name dan last name\n",
    "'''\n",
    "\n",
    "print(\"\\n\\n======= Spliting =======\")\n",
    "df = pd.DataFrame(\n",
    "    {\n",
    "        'Name':['Human A', 'Human B', 'Human C'],\n",
    "        'Age':[40, 34, 26]\n",
    "    }\n",
    ")\n",
    "\n",
    "print(\"Data Frame Awal\\n\", df)\n",
    "\n",
    "df[['First','Last']] = df.Name.str.split(expand=True) \n",
    "print(\"\\nData Frame setelah Split kolom 'Name', lalu menambahkan kedua kolom :\\n\", df)\n",
    "\n",
    "'''\n",
    "Interaction Future\n",
    "\n",
    "Mengkombinasikan dua atribut menjadi atribut yang lebih lebih baik.\n",
    "Contoh teknik Interaction Future :\n",
    "- feature selection, teknik lain dalam feature engineering.\n",
    "- menyusun (handcrafted)\n",
    "'''"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "718eff2d-4b73-43c4-a1a5-8ab6fad1e712",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}