{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "tags": [
     "parameters"
    ]
   },
   "outputs": [],
   "source": [
    "# notebook parameters\n",
    "\n",
    "import os\n",
    "import sys\n",
    "\n",
    "# replace this with your notebook working directory\n",
    "sys.path.append(\"/root/telco-churn-augmentation/\")\n",
    "\n",
    "\n",
    "spark_master = \"yarn\"\n",
    "app_name = \"churn-analytics\"\n",
    "input_prefix = \"\"\n",
    "input_file = \"churn-etl\"\n",
    "output_prefix = \"\"\n",
    "output_mode = \"overwrite\"\n",
    "output_kind = \"parquet\"\n",
    "input_kind = \"parquet\"\n",
    "driver_memory = '8g'\n",
    "executor_memory = '8g'\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pyspark\n",
    "\n",
    "session = pyspark.sql.SparkSession.builder \\\n",
    "    .master(spark_master) \\\n",
    "    .appName(app_name) \\\n",
    "    .config(\"spark.eventLog.enabled\", True) \\\n",
    "    .config(\"spark.eventLog.dir\", \".\") \\\n",
    "    .config(\"spark.driver.memory\", driver_memory) \\\n",
    "    .config(\"spark.executor.memory\", executor_memory) \\\n",
    "    .config(\"spark.rapids.memory.pinnedPool.size\", \"2G\") \\\n",
    "    .config(\"spark.sql.shuffle.partitions\", 16) \\\n",
    "    .config(\"spark.sql.files.maxPartitionBytes\", \"4096MB\") \\\n",
    "    .config(\"spark.rapids.sql.enabled\", True) \\\n",
    "    .config(\"spark.executor.cores\", 4) \\\n",
    "    .config(\"spark.task.cpus\", 1) \\\n",
    "    .config(\"spark.rapids.sql.concurrentGpuTasks\", 2) \\\n",
    "    .config(\"spark.task.resource.gpu.amount\", .5) \\\n",
    "    .config(\"spark.rapids.sql.variableFloatAgg.enabled\", True) \\\n",
    "    .config(\"spark.rapids.sql.explain\", \"NOT_ON_GPU\") \\\n",
    "    .config(\"spark.rapids.sql.decimalType.enabled\", \"True\") \\\n",
    "    .getOrCreate()\n",
    "session"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "import churn.eda\n",
    "import churn.etl\n",
    "\n",
    "churn.etl.register_options(\n",
    "    output_prefix = output_prefix,\n",
    "    output_mode = output_mode,\n",
    "    output_kind = output_kind,\n",
    "    input_kind = input_kind\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = churn.etl.read_df(session, input_prefix + input_file)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Generating reports"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "%%time\n",
    "\n",
    "summary = churn.eda.gen_summary(df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "session.catalog.listTables()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Slicing and dicing by tenure in quarters"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "grouped_by_quarters = session.table(\"cube_3\").select(\"3_month_spans\", \"Contract\", \"PaperlessBilling\", \"Churn\", \"Count\").toPandas()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "grouped_by_quarters = grouped_by_quarters.rename(columns = {'3_month_spans' : 'tenure_in_quarters'})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "import altair as alt\n",
    "alt.data_transformers.enable('json')\n",
    "\n",
    "alt.Chart(grouped_by_quarters.dropna()).mark_bar().encode(\n",
    "    x = 'tenure_in_quarters:O',\n",
    "    y = 'sum(Count):Q',\n",
    "    color = 'Churn:N',\n",
    "    column = 'Contract:N'\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "import altair as alt\n",
    "alt.data_transformers.enable('json')\n",
    "\n",
    "alt.Chart(grouped_by_quarters.dropna()).mark_bar().encode(\n",
    "    x = 'tenure_in_quarters:O',\n",
    "    y = 'sum(Count):Q',\n",
    "    column = 'Churn:N',\n",
    "    color = 'Contract:N'\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Finishing up"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# session.stop()"
   ]
  }
 ],
 "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": "3.8.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}