Referencing to columns when using PySpark

notebook.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "import findspark\n",
    "from pyspark.sql import SparkSession\n",
    "from pyspark.sql import functions as F\n",
    "findspark.init()\n",
    "\n",
    "\n",
    "spark = (\n",
    "    SparkSession.builder.appName(\"TestApp\")\n",
    "    .config(\"spark.driver.host\", \"localhost\")\n",
    "    .getOrCreate()\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Accessing Columns of a Dataframe"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "data = [(\"Alice\", 25, 3.14), (\"Bob\", 30, 2.71), (\"Charlie\", 35, 1.618)]\n",
    "df = spark.createDataFrame(data, [\"Name\", \"Age\", \"A nice number\"])\n",
    "df.toPandas()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### The 🐼 Pandas' way"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "df.toPandas().Age"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "df.toPandas()[\"A nice number\"]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### The ⚡️ Spark's Way"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "All the following calls are the same:\n",
    "\n",
    "```python\n",
    "df.select(\"Name\")\n",
    "df.select(df[\"Name\"])\n",
    "df.select(df.Name)\n",
    "df.select(F.col(\"Name\"))\n",
    "```\n",
    "\n",
    "For example:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "df.select(F.col(\"A nice number\")).toPandas()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Also the following have the same impact:\n",
    "\n",
    "```python\n",
    "df.withColumn(\"foo\", df.Age * 2)\n",
    "df.withColumn(\"foo\", df[\"Age\"] * 2)\n",
    "```"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "df.withColumn(\"foo\", df.Age * 2).toPandas()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Am I defined??\n",
    "\n",
    "![alt text](https://media.giphy.com/media/v1.Y2lkPTc5MGI3NjExYzI3aGtuOWVmNzlkenk4OG1kYXZpZjJjZmlkaWl6Ym9kNTh3dnhsYiZlcD12MV9naWZzX3RyZW5kaW5nJmN0PWc/7MDZS8zS1ixtJAUEul/giphy.gif \"I'm not defined!\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "#                                                   👇🏻👇🏻👇🏻👇🏻\n",
    "df.withColumn(\"foo\", df.Age * 2).withColumn(\"foo2\", \"foo\" * 2).toPandas()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "df.withColumn(\"foo\", df.Age * 2).withColumn(\"foo2\", F.col(\"foo\") * 2).toPandas()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "df2 = df.withColumn(\"foo\", df.Age * 2)\n",
    "# df2.withColumn(\"foo2\", df2.foo * 2).toPandas()\n",
    "df2.withColumn(\"foo2\", F.col(\"foo\") * 2).toPandas()"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": ".venv",
   "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": "3.12.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}