sparklyr_test.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "# Sparklyr workshop - R Users Galicia"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 16 July 2019"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Table of Contents\n",
    "\n",
    "- [What is sparklyr?](#What-is-sparklyr?)\n",
    "- [Kickstart](#Kickstart)\n",
    "- [R console](#Launch-R-console-first)\n",
    "- [Jupyter](#Launch-Jupyter)\n",
    "- [Uploading the data in JSON format](#Uploading-the-data-in-JSON-format)\n",
    "- [Spark and R in Jupyter](#Spark-and-R-in-Jupyter)\n",
    "- [Reading JSON into Spark context: `jscars.json`](#Reading-JSON-into-Spark-context:-jscars.json)\n",
    "    - [Basic data wrangling](#Basic-data-wrangling)\n",
    "    - [Using Hive functions and UDFs](#Using-Hive-functions-and-UDFs)\n",
    "    - [Pattern matching](#Pattern-matching-(PostgreSQL))\n",
    "    - [Graphs](#Graphs)\n",
    "    - [Models: K-means](#Models:-K-means)\n",
    "- [Kill YARN application and free memory](#Kill-YARN-application-and-free-memory)\n",
    "- [User Interface](#User-Interface)\n",
    "- [Session](#Session)\n",
    "    - [Paths recognised by R](#Paths-recognised-by-R)\n",
    "    - [R session info](#R-session-info)\n",
    "- [References](#References)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## What is sparklyr?\n",
    "\n",
    "- Exposes Spark's API (in Scala) from R.  \n",
    "\n",
    "- Spark lets us access to the Hadoop ecosystem.\n",
    "\n",
    "- To put it simple, <span style=\"color:#fffff;font-family:monospace;\"> dplyr</span> over Spark!.\n",
    "\n",
    "- Created by Rstudio in 2016.\n",
    "\n",
    "[How does Spark work?](https://aurora-mareviv.github.io/sparklyr_test/#/como-funciona-spark-2)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Kickstart"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Please, read the [pyspark course](https://github.com/javicacheiro/pyspark_course/blob/master/unit_1_tools.ipynb) here (points 1.1 to 1.3)."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "[Updated jul 2019]\n",
    "\n",
    "Open a terminal\n",
    "\n",
    "Connect to one of the login nodes of the Hadoop Hortonworks Data Platform (HDP) cluster (remember to activate Forticlient VPN first!):\n",
    "\n",
    "    MY_CESGA_USER='abcdef' # my CESGA username \n",
    "    ssh $MY_CESGA_USER@hadoop.cesga.es\n",
    "\n",
    "In the new HDP server you have to load the Anaconda module that has the R distribution that works with Spark. \n",
    "\n",
    "This command needs to be executed everytime a new console is opened at CESGA:\n",
    "\n",
    "    module load anaconda2  # run only once for every console \n",
    "    \n",
    "To see the available modules type:\n",
    "    \n",
    "    module available # module av"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Launch R console first"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### <font color='#b3b3b3'>Configuring Spark and R kernel using .bashrc (currently not recommended)</font>\n",
    "\n",
    "<br>\n",
    "<font color='#b3b3b3'>\n",
    "    \n",
    "*We could add environment variables to .bashrc to avoid setting `Sys.setenv` in `R`.*\n",
    "\n",
    "However, it is better to set `Sys.setenv()` as the **Anaconda version may change!** \n",
    "\n",
    "\n",
    "    cd $HOME\n",
    "    export R_PROFILE_USER=/usr/hdp/2.4.2.0-258/spark/R/lib/SparkR/profile/shell.R\n",
    "    # reload .bashrc\n",
    "    source .bashrc\n",
    "    \n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Open R\n",
    "\n",
    "Launch **R** (version 3.5.1 as of july 2019):\n",
    "\n",
    "    R\n",
    "\n",
    "### Install libraries in R\n",
    "\n",
    "Install the libraries as needed. When you finish, exit the interactive `R` session.\n",
    "\n",
    "    > install.packages(c(\"sparklyr\", \"dplyr\", \"knitr\", \"ggplot2\", \"repr\")\n",
    "    > q()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "'/usr/hdp/2.4.2.0-258/spark/R/lib/SparkR/profile/shell.R'"
      ],
      "text/latex": [
       "'/usr/hdp/2.4.2.0-258/spark/R/lib/SparkR/profile/shell.R'"
      ],
      "text/markdown": [
       "'/usr/hdp/2.4.2.0-258/spark/R/lib/SparkR/profile/shell.R'"
      ],
      "text/plain": [
       "[1] \"/usr/hdp/2.4.2.0-258/spark/R/lib/SparkR/profile/shell.R\""
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "Sys.getenv('R_PROFILE_USER')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Launch Jupyter\n",
    "\n",
    "To init a Jupyter notebook:\n",
    "\n",
    "- Open a new console window.\n",
    "\n",
    "- Connect to one of the login nodes of the Hadoop Hortonworks Data Platform (HDP) cluster: \n",
    "\n",
    "      ssh $MY_CESGA_USER@hadoop.cesga.es\n",
    "\n",
    "- Load Anaconda2 distribution:\n",
    "\n",
    "      module load anaconda2\n",
    "\n",
    "- Launch the Jupyter notebook: \n",
    "\n",
    "      start_jupyter\n",
    "    \n",
    "\n",
    "- Follow the instructions in the console. You may be asked to paste an URL token into your browser. \n",
    "\n",
    "- Open the provided link in the browser\n",
    "\n",
    "The notebook you have launched is tightly integrated with Spark in order to:\n",
    "\n",
    "- Automatically opens a R session with the cluster\n",
    "- Dynamically expands the required executors in the cluster\n",
    "- The console provides useful debugging information, showing Spark messages.\n",
    "\n",
    "When you finish your interactive work remember to close the notebook properly:\n",
    "\n",
    "- File -> Close and Halt\n",
    "\n",
    "\n",
    "**Tip**: it is a good idea to leave two consoles open simultaneously, one for Jupyter (to open the notebook) and the other for Bash and R commands.\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Uploading the data in JSON format "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "First, we need to upload our data from our NFS **HOME** to the HDFS **HOME** directory.  \n",
    "- You can add a single file, or an entire folder.\n",
    "- You can also list or delete files, folders.\n",
    "\n",
    "### HDFS overview and commands\n",
    "\n",
    "See the [pyspark course](https://github.com/javicacheiro/pyspark_course/blob/master/unit_1_tools.ipynb) here (point 1.2) for HDFS basic commands.\n",
    "\n",
    "Let's define some variables:"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "    # my username \n",
    "    MY_CESGA_USER='abcdef' \n",
    "    \n",
    "    # my HDFS HOME \n",
    "    MY_HDFS_HOME=/user/$MY_CESGA_USER/    \n",
    "    \n",
    "    # my NFS HOME \n",
    "    MY_NFS_HOME=$HOME"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "HDFS basic commands:"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "    # import data into hdfs (jscars.json): \n",
    "    hdfs dfs -put $MY_NFS_HOME/Rsession/sparklyr_start/data/jscars.json  # adapt to your path\n",
    "    \n",
    "    # list files \n",
    "    hdfs dfs -ls \n",
    "    # list files in a directory \n",
    "    hdfs dfs -ls $MY_HDFS_HOME\n",
    "    # create a directory called 'data': \n",
    "    hdfs dfs -mkdir data \n",
    "    # delete a directory: \n",
    "    hdfs dfs -rm -r -f data\n",
    "    # delete files, folders from hdfs\n",
    "    hdfs dfs -rm -r -f data/jscars.json"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Spark and R in Jupyter"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Set `R` environment variables (needed for Jupyter notebooks in the cluster `R` installation)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Sys.setenv(SPARK_HOME='/usr/hdp/2.4.2.0-258/spark/R/lib/SparkR/profile/shell.R') \n",
    "  # commented after being added to .bashrc\n",
    ".libPaths(c(file.path(Sys.getenv('SPARK_HOME'), 'R', 'lib'), .libPaths()))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "'/usr/hdp/2.4.2.0-258/spark'"
      ],
      "text/latex": [
       "'/usr/hdp/2.4.2.0-258/spark'"
      ],
      "text/markdown": [
       "'/usr/hdp/2.4.2.0-258/spark'"
      ],
      "text/plain": [
       "[1] \"/usr/hdp/2.4.2.0-258/spark\""
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "Sys.getenv('SPARK_HOME')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Load libraries:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "options(warn = -1) # disable warnings"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<ol>\n",
       "\t<li>TRUE</li>\n",
       "\t<li>TRUE</li>\n",
       "\t<li>TRUE</li>\n",
       "\t<li>TRUE</li>\n",
       "\t<li>TRUE</li>\n",
       "</ol>\n"
      ],
      "text/latex": [
       "\\begin{enumerate}\n",
       "\\item TRUE\n",
       "\\item TRUE\n",
       "\\item TRUE\n",
       "\\item TRUE\n",
       "\\item TRUE\n",
       "\\end{enumerate}\n"
      ],
      "text/markdown": [
       "1. TRUE\n",
       "2. TRUE\n",
       "3. TRUE\n",
       "4. TRUE\n",
       "5. TRUE\n",
       "\n",
       "\n"
      ],
      "text/plain": [
       "[[1]]\n",
       "[1] TRUE\n",
       "\n",
       "[[2]]\n",
       "[1] TRUE\n",
       "\n",
       "[[3]]\n",
       "[1] TRUE\n",
       "\n",
       "[[4]]\n",
       "[1] TRUE\n",
       "\n",
       "[[5]]\n",
       "[1] TRUE\n"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "x <- c(\"sparklyr\", \"dplyr\", \"knitr\", \"ggplot2\", \"repr\")\n",
    "lapply(x, require, character.only = TRUE, quietly = TRUE)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Loading spark context (RDDs)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Currently there are three types of contexts:\n",
    "\n",
    "- Local context: \n",
    "    - Interactive. \n",
    "    - If the user exits session, the tasks are terminated (use `screen` to run after session close).\n",
    "    - All processes reside in the LOGIN node (both drivers and executors).\n",
    "    - Can only be used for tasks that require very few resources.\n",
    "    - .Rmd notebook to run in your local computer\n",
    "- YARN-client: \n",
    "    - Interactive.\n",
    "    - If the user exits session, the tasks are terminated (use `screen` to run after session close).\n",
    "    - The driver resides in the LOGIN node, but the executors are in the CLUSTER nodes. Thus, executors can use all the memory available for the task in the CLUSTER nodes.\n",
    "    - Can be used for memory-intensive tasks.\n",
    "- YARN-cluster:\n",
    "    - Not interactive.\n",
    "    - Both the driver and the executors reside in the CLUSTER nodes.\n",
    "    - Can be used for memory-intensive tasks.\n",
    "    - Currenty doesn't seem available for this version of `R/sparklyr`.\n",
    "    \n",
    "Defining a new context (`sc`) overwrites the previous one."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "SPARK_HOME = Sys.getenv('SPARK_HOME')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Re-using existing Spark connection to yarn-client\n"
     ]
    }
   ],
   "source": [
    "# Initiating spark context: local (for 'low memory' tasks only!)\n",
    "# sc <- spark_connect(master = \"local\", spark_home = SPARK_HOME)\n",
    "\n",
    "# Initiating spark context: yarn (for loading bigger datasets)\n",
    "sc <- spark_connect(master = \"yarn-client\", spark_home = SPARK_HOME)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Reading JSON into Spark context: `jscars.json`"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The data `jscars.json` is the `mtcars` dataset saved in a JSON format.\n",
    "\n",
    "Beware: the command `getwd()` gives us our NFS working directory, not an HDFS directory."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "wdir <- getwd() # NFS working directory, not HDFS!"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "To import files, we must refer to the path in the HDFS directory tree (see `$ hdfs dfs -ls`):"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "jscars <- spark_read_json(sc, name = \"jscars\", path = \"jscars.json\") # HDFS working directory"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<table>\n",
       "<thead><tr><th scope=col>am</th><th scope=col>carb</th><th scope=col>cyl</th><th scope=col>disp</th><th scope=col>drat</th><th scope=col>gear</th><th scope=col>hp</th><th scope=col>mpg</th><th scope=col>qsec</th><th scope=col>vs</th><th scope=col>wt</th></tr></thead>\n",
       "<tbody>\n",
       "\t<tr><td>1    </td><td>4    </td><td>6    </td><td>160  </td><td>3.90 </td><td>4    </td><td>110  </td><td>21.0 </td><td>16.46</td><td>0    </td><td>2.620</td></tr>\n",
       "\t<tr><td>1    </td><td>4    </td><td>6    </td><td>160  </td><td>3.90 </td><td>4    </td><td>110  </td><td>21.0 </td><td>17.02</td><td>0    </td><td>2.875</td></tr>\n",
       "\t<tr><td>1    </td><td>1    </td><td>4    </td><td>108  </td><td>3.85 </td><td>4    </td><td> 93  </td><td>22.8 </td><td>18.61</td><td>1    </td><td>2.320</td></tr>\n",
       "\t<tr><td>0    </td><td>1    </td><td>6    </td><td>258  </td><td>3.08 </td><td>3    </td><td>110  </td><td>21.4 </td><td>19.44</td><td>1    </td><td>3.215</td></tr>\n",
       "\t<tr><td>0    </td><td>2    </td><td>8    </td><td>360  </td><td>3.15 </td><td>3    </td><td>175  </td><td>18.7 </td><td>17.02</td><td>0    </td><td>3.440</td></tr>\n",
       "\t<tr><td>0    </td><td>1    </td><td>6    </td><td>225  </td><td>2.76 </td><td>3    </td><td>105  </td><td>18.1 </td><td>20.22</td><td>1    </td><td>3.460</td></tr>\n",
       "</tbody>\n",
       "</table>\n"
      ],
      "text/latex": [
       "\\begin{tabular}{r|lllllllllll}\n",
       " am & carb & cyl & disp & drat & gear & hp & mpg & qsec & vs & wt\\\\\n",
       "\\hline\n",
       "\t 1     & 4     & 6     & 160   & 3.90  & 4     & 110   & 21.0  & 16.46 & 0     & 2.620\\\\\n",
       "\t 1     & 4     & 6     & 160   & 3.90  & 4     & 110   & 21.0  & 17.02 & 0     & 2.875\\\\\n",
       "\t 1     & 1     & 4     & 108   & 3.85  & 4     &  93   & 22.8  & 18.61 & 1     & 2.320\\\\\n",
       "\t 0     & 1     & 6     & 258   & 3.08  & 3     & 110   & 21.4  & 19.44 & 1     & 3.215\\\\\n",
       "\t 0     & 2     & 8     & 360   & 3.15  & 3     & 175   & 18.7  & 17.02 & 0     & 3.440\\\\\n",
       "\t 0     & 1     & 6     & 225   & 2.76  & 3     & 105   & 18.1  & 20.22 & 1     & 3.460\\\\\n",
       "\\end{tabular}\n"
      ],
      "text/markdown": [
       "\n",
       "am | carb | cyl | disp | drat | gear | hp | mpg | qsec | vs | wt | \n",
       "|---|---|---|---|---|---|\n",
       "| 1     | 4     | 6     | 160   | 3.90  | 4     | 110   | 21.0  | 16.46 | 0     | 2.620 | \n",
       "| 1     | 4     | 6     | 160   | 3.90  | 4     | 110   | 21.0  | 17.02 | 0     | 2.875 | \n",
       "| 1     | 1     | 4     | 108   | 3.85  | 4     |  93   | 22.8  | 18.61 | 1     | 2.320 | \n",
       "| 0     | 1     | 6     | 258   | 3.08  | 3     | 110   | 21.4  | 19.44 | 1     | 3.215 | \n",
       "| 0     | 2     | 8     | 360   | 3.15  | 3     | 175   | 18.7  | 17.02 | 0     | 3.440 | \n",
       "| 0     | 1     | 6     | 225   | 2.76  | 3     | 105   | 18.1  | 20.22 | 1     | 3.460 | \n",
       "\n",
       "\n"
      ],
      "text/plain": [
       "  am carb cyl disp drat gear hp  mpg  qsec  vs wt   \n",
       "1 1  4    6   160  3.90 4    110 21.0 16.46 0  2.620\n",
       "2 1  4    6   160  3.90 4    110 21.0 17.02 0  2.875\n",
       "3 1  1    4   108  3.85 4     93 22.8 18.61 1  2.320\n",
       "4 0  1    6   258  3.08 3    110 21.4 19.44 1  3.215\n",
       "5 0  2    8   360  3.15 3    175 18.7 17.02 0  3.440\n",
       "6 0  1    6   225  2.76 3    105 18.1 20.22 1  3.460"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "jscars %>%\n",
    "  head(6) %>%\n",
    "  collect()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Basic data wrangling"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can make good use of `magrittr::%>%` *pipes* for data wrangling or window operations:  \n",
    "The `warnings` issued here are usually not important. They depend on the `sparklyr` version. See [here](https://github.com/rstudio/sparklyr/issues/426) for more info."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<table>\n",
       "<thead><tr><th scope=col>vs</th><th scope=col>gear</th><th scope=col>mpg_rank</th><th scope=col>horsepower_by_gear</th></tr></thead>\n",
       "<tbody>\n",
       "\t<tr><td>0</td><td>3</td><td>1</td><td>4</td></tr>\n",
       "\t<tr><td>0</td><td>3</td><td>1</td><td>5</td></tr>\n",
       "\t<tr><td>0</td><td>3</td><td>3</td><td>1</td></tr>\n",
       "\t<tr><td>0</td><td>3</td><td>4</td><td>1</td></tr>\n",
       "\t<tr><td>0</td><td>3</td><td>5</td><td>3</td></tr>\n",
       "\t<tr><td>0</td><td>3</td><td>6</td><td>6</td></tr>\n",
       "</tbody>\n",
       "</table>\n"
      ],
      "text/latex": [
       "\\begin{tabular}{r|llll}\n",
       " vs & gear & mpg\\_rank & horsepower\\_by\\_gear\\\\\n",
       "\\hline\n",
       "\t 0 & 3 & 1 & 4\\\\\n",
       "\t 0 & 3 & 1 & 5\\\\\n",
       "\t 0 & 3 & 3 & 1\\\\\n",
       "\t 0 & 3 & 4 & 1\\\\\n",
       "\t 0 & 3 & 5 & 3\\\\\n",
       "\t 0 & 3 & 6 & 6\\\\\n",
       "\\end{tabular}\n"
      ],
      "text/markdown": [
       "\n",
       "vs | gear | mpg_rank | horsepower_by_gear | \n",
       "|---|---|---|---|---|---|\n",
       "| 0 | 3 | 1 | 4 | \n",
       "| 0 | 3 | 1 | 5 | \n",
       "| 0 | 3 | 3 | 1 | \n",
       "| 0 | 3 | 4 | 1 | \n",
       "| 0 | 3 | 5 | 3 | \n",
       "| 0 | 3 | 6 | 6 | \n",
       "\n",
       "\n"
      ],
      "text/plain": [
       "  vs gear mpg_rank horsepower_by_gear\n",
       "1 0  3    1        4                 \n",
       "2 0  3    1        5                 \n",
       "3 0  3    3        1                 \n",
       "4 0  3    4        1                 \n",
       "5 0  3    5        3                 \n",
       "6 0  3    6        6                 "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "jscars %>%\n",
    "  group_by(vs) %>%\n",
    "  filter(gear == 3, hp > 100) %>%\n",
    "  mutate(horsepower_by_gear = rank(desc(hp / gear))) %>%\n",
    "  mutate(mpg_rank = rank(mpg)) %>%\n",
    "  select(gear, mpg_rank, horsepower_by_gear) %>% \n",
    "  head(6) %>%\n",
    "  collect()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Using Hive functions and UDFs\n",
    "\n",
    "Fortunately, `sparklyr` integrates well with other components in the Hadoop-Spark stack. Here, an example with Hive functions."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<table>\n",
       "<thead><tr><th scope=col>am</th><th scope=col>carb</th><th scope=col>cyl</th><th scope=col>disp</th><th scope=col>drat</th><th scope=col>gear</th><th scope=col>hp</th><th scope=col>mpg</th><th scope=col>qsec</th><th scope=col>vs</th><th scope=col>wt</th><th scope=col>date_today</th><th scope=col>date_last</th><th scope=col>days_past</th></tr></thead>\n",
       "<tbody>\n",
       "\t<tr><td>1         </td><td>4         </td><td>6         </td><td>160       </td><td>3,90      </td><td>4         </td><td>110       </td><td>21,0      </td><td>16,46     </td><td>0         </td><td>2,620     </td><td>2019-07-15</td><td>2019-01-01</td><td>196       </td></tr>\n",
       "\t<tr><td>1         </td><td>4         </td><td>6         </td><td>160       </td><td>3,90      </td><td>4         </td><td>110       </td><td>21,0      </td><td>17,02     </td><td>0         </td><td>2,875     </td><td>2019-07-15</td><td>2019-01-01</td><td>196       </td></tr>\n",
       "\t<tr><td>1         </td><td>1         </td><td>4         </td><td>108       </td><td>3,85      </td><td>4         </td><td> 93       </td><td>22,8      </td><td>18,61     </td><td>1         </td><td>2,320     </td><td>2019-07-15</td><td>2019-01-01</td><td>196       </td></tr>\n",
       "\t<tr><td>0         </td><td>1         </td><td>6         </td><td>258       </td><td>3,08      </td><td>3         </td><td>110       </td><td>21,4      </td><td>19,44     </td><td>1         </td><td>3,215     </td><td>2019-07-15</td><td>2019-01-01</td><td>196       </td></tr>\n",
       "\t<tr><td>0         </td><td>2         </td><td>8         </td><td>360       </td><td>3,15      </td><td>3         </td><td>175       </td><td>18,7      </td><td>17,02     </td><td>0         </td><td>3,440     </td><td>2019-07-15</td><td>2019-01-01</td><td>196       </td></tr>\n",
       "\t<tr><td>0         </td><td>1         </td><td>6         </td><td>225       </td><td>2,76      </td><td>3         </td><td>105       </td><td>18,1      </td><td>20,22     </td><td>1         </td><td>3,460     </td><td>2019-07-15</td><td>2019-01-01</td><td>196       </td></tr>\n",
       "</tbody>\n",
       "</table>\n"
      ],
      "text/latex": [
       "\\begin{tabular}{r|llllllllllllll}\n",
       " am & carb & cyl & disp & drat & gear & hp & mpg & qsec & vs & wt & date\\_today & date\\_last & days\\_past\\\\\n",
       "\\hline\n",
       "\t 1          & 4          & 6          & 160        & 3,90       & 4          & 110        & 21,0       & 16,46      & 0          & 2,620      & 2019-07-15 & 2019-01-01 & 196       \\\\\n",
       "\t 1          & 4          & 6          & 160        & 3,90       & 4          & 110        & 21,0       & 17,02      & 0          & 2,875      & 2019-07-15 & 2019-01-01 & 196       \\\\\n",
       "\t 1          & 1          & 4          & 108        & 3,85       & 4          &  93        & 22,8       & 18,61      & 1          & 2,320      & 2019-07-15 & 2019-01-01 & 196       \\\\\n",
       "\t 0          & 1          & 6          & 258        & 3,08       & 3          & 110        & 21,4       & 19,44      & 1          & 3,215      & 2019-07-15 & 2019-01-01 & 196       \\\\\n",
       "\t 0          & 2          & 8          & 360        & 3,15       & 3          & 175        & 18,7       & 17,02      & 0          & 3,440      & 2019-07-15 & 2019-01-01 & 196       \\\\\n",
       "\t 0          & 1          & 6          & 225        & 2,76       & 3          & 105        & 18,1       & 20,22      & 1          & 3,460      & 2019-07-15 & 2019-01-01 & 196       \\\\\n",
       "\\end{tabular}\n"
      ],
      "text/markdown": [
       "\n",
       "am | carb | cyl | disp | drat | gear | hp | mpg | qsec | vs | wt | date_today | date_last | days_past | \n",
       "|---|---|---|---|---|---|\n",
       "| 1          | 4          | 6          | 160        | 3,90       | 4          | 110        | 21,0       | 16,46      | 0          | 2,620      | 2019-07-15 | 2019-01-01 | 196        | \n",
       "| 1          | 4          | 6          | 160        | 3,90       | 4          | 110        | 21,0       | 17,02      | 0          | 2,875      | 2019-07-15 | 2019-01-01 | 196        | \n",
       "| 1          | 1          | 4          | 108        | 3,85       | 4          |  93        | 22,8       | 18,61      | 1          | 2,320      | 2019-07-15 | 2019-01-01 | 196        | \n",
       "| 0          | 1          | 6          | 258        | 3,08       | 3          | 110        | 21,4       | 19,44      | 1          | 3,215      | 2019-07-15 | 2019-01-01 | 196        | \n",
       "| 0          | 2          | 8          | 360        | 3,15       | 3          | 175        | 18,7       | 17,02      | 0          | 3,440      | 2019-07-15 | 2019-01-01 | 196        | \n",
       "| 0          | 1          | 6          | 225        | 2,76       | 3          | 105        | 18,1       | 20,22      | 1          | 3,460      | 2019-07-15 | 2019-01-01 | 196        | \n",
       "\n",
       "\n"
      ],
      "text/plain": [
       "  am carb cyl disp drat gear hp  mpg  qsec  vs wt    date_today date_last \n",
       "1 1  4    6   160  3,90 4    110 21,0 16,46 0  2,620 2019-07-15 2019-01-01\n",
       "2 1  4    6   160  3,90 4    110 21,0 17,02 0  2,875 2019-07-15 2019-01-01\n",
       "3 1  1    4   108  3,85 4     93 22,8 18,61 1  2,320 2019-07-15 2019-01-01\n",
       "4 0  1    6   258  3,08 3    110 21,4 19,44 1  3,215 2019-07-15 2019-01-01\n",
       "5 0  2    8   360  3,15 3    175 18,7 17,02 0  3,440 2019-07-15 2019-01-01\n",
       "6 0  1    6   225  2,76 3    105 18,1 20,22 1  3,460 2019-07-15 2019-01-01\n",
       "  days_past\n",
       "1 196      \n",
       "2 196      \n",
       "3 196      \n",
       "4 196      \n",
       "5 196      \n",
       "6 196      "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "jscars_dates <- jscars %>%\n",
    "  mutate(\n",
    "      date_today = as.Date( current_date() ) # current date retrieved by Hive function \"current_date\"\n",
    "      ) %>%\n",
    "  mutate(\n",
    "      date_last = \"2019-01-01\" \n",
    "      ) %>%\n",
    "  mutate(\n",
    "      days_past = datediff(date_today, date_last) # days substraction using Hive function \"datediff\"\n",
    "      ) \n",
    "\n",
    "jscars_dates %>%\n",
    "head() %>%\n",
    "collect()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Pattern matching (PostgreSQL)\n",
    "We also have the possibility to make regular expressions using the PostgreSQL API, available to sparklyr:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<table>\n",
       "<thead><tr><th scope=col>am</th><th scope=col>carb</th><th scope=col>cyl</th><th scope=col>disp</th><th scope=col>drat</th><th scope=col>gear</th><th scope=col>hp</th><th scope=col>mpg</th><th scope=col>qsec</th><th scope=col>vs</th><th scope=col>wt</th><th scope=col>date_today</th><th scope=col>date_last</th><th scope=col>days_past</th><th scope=col>replaced_year</th></tr></thead>\n",
       "<tbody>\n",
       "\t<tr><td>1         </td><td>4         </td><td>6         </td><td>160       </td><td>3,90      </td><td>4         </td><td>110       </td><td>21,0      </td><td>16,46     </td><td>0         </td><td>2,620     </td><td>2019-07-15</td><td>2019-01-01</td><td>196       </td><td>9999-01-01</td></tr>\n",
       "\t<tr><td>1         </td><td>4         </td><td>6         </td><td>160       </td><td>3,90      </td><td>4         </td><td>110       </td><td>21,0      </td><td>17,02     </td><td>0         </td><td>2,875     </td><td>2019-07-15</td><td>2019-01-01</td><td>196       </td><td>9999-01-01</td></tr>\n",
       "\t<tr><td>1         </td><td>1         </td><td>4         </td><td>108       </td><td>3,85      </td><td>4         </td><td> 93       </td><td>22,8      </td><td>18,61     </td><td>1         </td><td>2,320     </td><td>2019-07-15</td><td>2019-01-01</td><td>196       </td><td>9999-01-01</td></tr>\n",
       "\t<tr><td>0         </td><td>1         </td><td>6         </td><td>258       </td><td>3,08      </td><td>3         </td><td>110       </td><td>21,4      </td><td>19,44     </td><td>1         </td><td>3,215     </td><td>2019-07-15</td><td>2019-01-01</td><td>196       </td><td>9999-01-01</td></tr>\n",
       "\t<tr><td>0         </td><td>2         </td><td>8         </td><td>360       </td><td>3,15      </td><td>3         </td><td>175       </td><td>18,7      </td><td>17,02     </td><td>0         </td><td>3,440     </td><td>2019-07-15</td><td>2019-01-01</td><td>196       </td><td>9999-01-01</td></tr>\n",
       "\t<tr><td>0         </td><td>1         </td><td>6         </td><td>225       </td><td>2,76      </td><td>3         </td><td>105       </td><td>18,1      </td><td>20,22     </td><td>1         </td><td>3,460     </td><td>2019-07-15</td><td>2019-01-01</td><td>196       </td><td>9999-01-01</td></tr>\n",
       "</tbody>\n",
       "</table>\n"
      ],
      "text/latex": [
       "\\begin{tabular}{r|lllllllllllllll}\n",
       " am & carb & cyl & disp & drat & gear & hp & mpg & qsec & vs & wt & date\\_today & date\\_last & days\\_past & replaced\\_year\\\\\n",
       "\\hline\n",
       "\t 1          & 4          & 6          & 160        & 3,90       & 4          & 110        & 21,0       & 16,46      & 0          & 2,620      & 2019-07-15 & 2019-01-01 & 196        & 9999-01-01\\\\\n",
       "\t 1          & 4          & 6          & 160        & 3,90       & 4          & 110        & 21,0       & 17,02      & 0          & 2,875      & 2019-07-15 & 2019-01-01 & 196        & 9999-01-01\\\\\n",
       "\t 1          & 1          & 4          & 108        & 3,85       & 4          &  93        & 22,8       & 18,61      & 1          & 2,320      & 2019-07-15 & 2019-01-01 & 196        & 9999-01-01\\\\\n",
       "\t 0          & 1          & 6          & 258        & 3,08       & 3          & 110        & 21,4       & 19,44      & 1          & 3,215      & 2019-07-15 & 2019-01-01 & 196        & 9999-01-01\\\\\n",
       "\t 0          & 2          & 8          & 360        & 3,15       & 3          & 175        & 18,7       & 17,02      & 0          & 3,440      & 2019-07-15 & 2019-01-01 & 196        & 9999-01-01\\\\\n",
       "\t 0          & 1          & 6          & 225        & 2,76       & 3          & 105        & 18,1       & 20,22      & 1          & 3,460      & 2019-07-15 & 2019-01-01 & 196        & 9999-01-01\\\\\n",
       "\\end{tabular}\n"
      ],
      "text/markdown": [
       "\n",
       "am | carb | cyl | disp | drat | gear | hp | mpg | qsec | vs | wt | date_today | date_last | days_past | replaced_year | \n",
       "|---|---|---|---|---|---|\n",
       "| 1          | 4          | 6          | 160        | 3,90       | 4          | 110        | 21,0       | 16,46      | 0          | 2,620      | 2019-07-15 | 2019-01-01 | 196        | 9999-01-01 | \n",
       "| 1          | 4          | 6          | 160        | 3,90       | 4          | 110        | 21,0       | 17,02      | 0          | 2,875      | 2019-07-15 | 2019-01-01 | 196        | 9999-01-01 | \n",
       "| 1          | 1          | 4          | 108        | 3,85       | 4          |  93        | 22,8       | 18,61      | 1          | 2,320      | 2019-07-15 | 2019-01-01 | 196        | 9999-01-01 | \n",
       "| 0          | 1          | 6          | 258        | 3,08       | 3          | 110        | 21,4       | 19,44      | 1          | 3,215      | 2019-07-15 | 2019-01-01 | 196        | 9999-01-01 | \n",
       "| 0          | 2          | 8          | 360        | 3,15       | 3          | 175        | 18,7       | 17,02      | 0          | 3,440      | 2019-07-15 | 2019-01-01 | 196        | 9999-01-01 | \n",
       "| 0          | 1          | 6          | 225        | 2,76       | 3          | 105        | 18,1       | 20,22      | 1          | 3,460      | 2019-07-15 | 2019-01-01 | 196        | 9999-01-01 | \n",
       "\n",
       "\n"
      ],
      "text/plain": [
       "  am carb cyl disp drat gear hp  mpg  qsec  vs wt    date_today date_last \n",
       "1 1  4    6   160  3,90 4    110 21,0 16,46 0  2,620 2019-07-15 2019-01-01\n",
       "2 1  4    6   160  3,90 4    110 21,0 17,02 0  2,875 2019-07-15 2019-01-01\n",
       "3 1  1    4   108  3,85 4     93 22,8 18,61 1  2,320 2019-07-15 2019-01-01\n",
       "4 0  1    6   258  3,08 3    110 21,4 19,44 1  3,215 2019-07-15 2019-01-01\n",
       "5 0  2    8   360  3,15 3    175 18,7 17,02 0  3,440 2019-07-15 2019-01-01\n",
       "6 0  1    6   225  2,76 3    105 18,1 20,22 1  3,460 2019-07-15 2019-01-01\n",
       "  days_past replaced_year\n",
       "1 196       9999-01-01   \n",
       "2 196       9999-01-01   \n",
       "3 196       9999-01-01   \n",
       "4 196       9999-01-01   \n",
       "5 196       9999-01-01   \n",
       "6 196       9999-01-01   "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "jscars_dates %>%\n",
    "  mutate(\n",
    "      replaced_year = regexp_replace(date_last, '2019', '9999') # PostgreSQL regexp\n",
    "      ) %>%\n",
    "  head() %>%\n",
    "  collect()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Graphs "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "`geom_smooth()` using method = 'loess' and formula 'y ~ x'\n"
     ]
    },
    {
     "data": {},
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAtAAAAHgCAIAAAADp837AAAACXBIWXMAABJ0AAASdAHeZh94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",
      "text/plain": [
       "plot without title"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "options(repr.plot.width=6, repr.plot.height=4) #repr library\n",
    "\n",
    "jscars %>%\n",
    "  collect() %>%\n",
    "    ggplot(aes(wt, mpg)) +\n",
    "    geom_point(aes(colour=factor(gear), size=hp)) + \n",
    "    geom_smooth(colour=\"#737373\", alpha=0.3) + \n",
    "    theme_bw()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Models: K-means "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "kmeans_model <- jscars %>%\n",
    "  select(wt, mpg) %>%\n",
    "  ml_kmeans(y ~ wt + mpg, k = 3)\n",
    "\n",
    "#print(kmeans_model)\n",
    "# kmeans_model %>% str()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<table>\n",
       "<thead><tr><th scope=col>am</th><th scope=col>carb</th><th scope=col>cyl</th><th scope=col>disp</th><th scope=col>drat</th><th scope=col>gear</th><th scope=col>hp</th><th scope=col>mpg</th><th scope=col>qsec</th><th scope=col>vs</th><th scope=col>wt</th><th scope=col>features</th><th scope=col>prediction</th></tr></thead>\n",
       "<tbody>\n",
       "\t<tr><td>1          </td><td>4          </td><td>6          </td><td>160        </td><td>3,90       </td><td>4          </td><td>110        </td><td>21,0       </td><td>16,46      </td><td>0          </td><td>2,620      </td><td>2,62, 21,00</td><td>2          </td></tr>\n",
       "\t<tr><td>1            </td><td>4            </td><td>6            </td><td>160          </td><td>3,90         </td><td>4            </td><td>110          </td><td>21,0         </td><td>17,02        </td><td>0            </td><td>2,875        </td><td>2,875, 21,000</td><td>2            </td></tr>\n",
       "\t<tr><td>1          </td><td>1          </td><td>4          </td><td>108        </td><td>3,85       </td><td>4          </td><td> 93        </td><td>22,8       </td><td>18,61      </td><td>1          </td><td>2,320      </td><td>2,32, 22,80</td><td>2          </td></tr>\n",
       "\t<tr><td>0            </td><td>1            </td><td>6            </td><td>258          </td><td>3,08         </td><td>3            </td><td>110          </td><td>21,4         </td><td>19,44        </td><td>1            </td><td>3,215        </td><td>3,215, 21,400</td><td>2            </td></tr>\n",
       "\t<tr><td>0          </td><td>2          </td><td>8          </td><td>360        </td><td>3,15       </td><td>3          </td><td>175        </td><td>18,7       </td><td>17,02      </td><td>0          </td><td>3,440      </td><td>3,44, 18,70</td><td>2          </td></tr>\n",
       "\t<tr><td>0          </td><td>1          </td><td>6          </td><td>225        </td><td>2,76       </td><td>3          </td><td>105        </td><td>18,1       </td><td>20,22      </td><td>1          </td><td>3,460      </td><td>3,46, 18,10</td><td>2          </td></tr>\n",
       "</tbody>\n",
       "</table>\n"
      ],
      "text/latex": [
       "\\begin{tabular}{r|lllllllllllll}\n",
       " am & carb & cyl & disp & drat & gear & hp & mpg & qsec & vs & wt & features & prediction\\\\\n",
       "\\hline\n",
       "\t 1           & 4           & 6           & 160         & 3,90        & 4           & 110         & 21,0        & 16,46       & 0           & 2,620       & 2,62, 21,00 & 2          \\\\\n",
       "\t 1             & 4             & 6             & 160           & 3,90          & 4             & 110           & 21,0          & 17,02         & 0             & 2,875         & 2,875, 21,000 & 2            \\\\\n",
       "\t 1           & 1           & 4           & 108         & 3,85        & 4           &  93         & 22,8        & 18,61       & 1           & 2,320       & 2,32, 22,80 & 2          \\\\\n",
       "\t 0             & 1             & 6             & 258           & 3,08          & 3             & 110           & 21,4          & 19,44         & 1             & 3,215         & 3,215, 21,400 & 2            \\\\\n",
       "\t 0           & 2           & 8           & 360         & 3,15        & 3           & 175         & 18,7        & 17,02       & 0           & 3,440       & 3,44, 18,70 & 2          \\\\\n",
       "\t 0           & 1           & 6           & 225         & 2,76        & 3           & 105         & 18,1        & 20,22       & 1           & 3,460       & 3,46, 18,10 & 2          \\\\\n",
       "\\end{tabular}\n"
      ],
      "text/markdown": [
       "\n",
       "am | carb | cyl | disp | drat | gear | hp | mpg | qsec | vs | wt | features | prediction | \n",
       "|---|---|---|---|---|---|\n",
       "| 1           | 4           | 6           | 160         | 3,90        | 4           | 110         | 21,0        | 16,46       | 0           | 2,620       | 2,62, 21,00 | 2           | \n",
       "| 1             | 4             | 6             | 160           | 3,90          | 4             | 110           | 21,0          | 17,02         | 0             | 2,875         | 2,875, 21,000 | 2             | \n",
       "| 1           | 1           | 4           | 108         | 3,85        | 4           |  93         | 22,8        | 18,61       | 1           | 2,320       | 2,32, 22,80 | 2           | \n",
       "| 0             | 1             | 6             | 258           | 3,08          | 3             | 110           | 21,4          | 19,44         | 1             | 3,215         | 3,215, 21,400 | 2             | \n",
       "| 0           | 2           | 8           | 360         | 3,15        | 3           | 175         | 18,7        | 17,02       | 0           | 3,440       | 3,44, 18,70 | 2           | \n",
       "| 0           | 1           | 6           | 225         | 2,76        | 3           | 105         | 18,1        | 20,22       | 1           | 3,460       | 3,46, 18,10 | 2           | \n",
       "\n",
       "\n"
      ],
      "text/plain": [
       "  am carb cyl disp drat gear hp  mpg  qsec  vs wt    features      prediction\n",
       "1 1  4    6   160  3,90 4    110 21,0 16,46 0  2,620 2,62, 21,00   2         \n",
       "2 1  4    6   160  3,90 4    110 21,0 17,02 0  2,875 2,875, 21,000 2         \n",
       "3 1  1    4   108  3,85 4     93 22,8 18,61 1  2,320 2,32, 22,80   2         \n",
       "4 0  1    6   258  3,08 3    110 21,4 19,44 1  3,215 3,215, 21,400 2         \n",
       "5 0  2    8   360  3,15 3    175 18,7 17,02 0  3,440 3,44, 18,70   2         \n",
       "6 0  1    6   225  2,76 3    105 18,1 20,22 1  3,460 3,46, 18,10   2         "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "   \n",
       "     0  1  2\n",
       "  0 10  0  9\n",
       "  1  2  6  5"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# predict the associated class\n",
    "predicts <- sdf_predict(jscars, kmeans_model) \n",
    "predicted <- collect(predicts)\n",
    "collect(head(predicted))\n",
    "base::table(predicted$am, predicted$prediction)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {},
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAtAAAAHgCAIAAAADp837AAAACXBIWXMAABJ0AAASdAHeZh94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\nRK5CYII=",
      "text/plain": [
       "plot without title"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# plot cluster membership\n",
    "sdf_predict(jscars, kmeans_model) %>%\n",
    "  collect() %>%\n",
    "  ggplot(aes(wt, mpg)) +\n",
    "    geom_point(aes(wt, mpg, col = factor(prediction + 1)),\n",
    "               size = 2, alpha = 0.5) + \n",
    "    geom_point(data = kmeans_model$centers, aes(wt, mpg),\n",
    "               col = scales::muted(c(\"red\", \"green\", \"blue\")),\n",
    "               pch = 'x', size = 12) +\n",
    "    scale_color_discrete(name = \"Predicted Cluster\",\n",
    "                         labels = paste(\"Cluster\", 1:3)) +\n",
    "    labs(\n",
    "      x = \"wt\",\n",
    "      y = \"mpg\",\n",
    "      title = \"K-Means Clustering\",\n",
    "      subtitle = \"Use Spark.ML to predict cluster membership with the jscars dataset.\"\n",
    "    ) +\n",
    "    theme_bw()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Kill YARN application and free memory\n",
    "\n",
    "How to see the aggregated logs produced by a job:\n",
    "      \n",
    "    yarn logs -applicationId <applicationID> | less\n",
    "\n",
    "How to see the list of running jobs:\n",
    "    \n",
    "    yarn application -list\n",
    "\n",
    "Kill an application:\n",
    "    \n",
    "    yarn application -kill <applicationID>\n",
    "    \n",
    "For further information check [The YARN Tutorial](http://bigdata.cesga.es/tutorials/yarn.html#/)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## User Interface\n",
    "\n",
    "Login to have access to the Spark UI\n",
    "\n",
    "[https://bigdata.cesga.es/web-ui/app](https://bigdata.cesga.es/web-ui/app/#/login)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Session "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Paths recognised by R "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<ol class=list-inline>\n",
       "\t<li>'/usr/hdp/2.4.2.0-258/spark/R/lib'</li>\n",
       "\t<li>'/home/chu/ar/mbg/R/libraries/3.5'</li>\n",
       "\t<li>'/opt/cesga/anaconda/Anaconda2-5.3.0/lib/R/library'</li>\n",
       "</ol>\n"
      ],
      "text/latex": [
       "\\begin{enumerate*}\n",
       "\\item '/usr/hdp/2.4.2.0-258/spark/R/lib'\n",
       "\\item '/home/chu/ar/mbg/R/libraries/3.5'\n",
       "\\item '/opt/cesga/anaconda/Anaconda2-5.3.0/lib/R/library'\n",
       "\\end{enumerate*}\n"
      ],
      "text/markdown": [
       "1. '/usr/hdp/2.4.2.0-258/spark/R/lib'\n",
       "2. '/home/chu/ar/mbg/R/libraries/3.5'\n",
       "3. '/opt/cesga/anaconda/Anaconda2-5.3.0/lib/R/library'\n",
       "\n",
       "\n"
      ],
      "text/plain": [
       "[1] \"/usr/hdp/2.4.2.0-258/spark/R/lib\"                 \n",
       "[2] \"/home/chu/ar/mbg/R/libraries/3.5\"                 \n",
       "[3] \"/opt/cesga/anaconda/Anaconda2-5.3.0/lib/R/library\""
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "'/opt/cesga/anaconda/Anaconda2-5.3.0/lib/R'"
      ],
      "text/latex": [
       "'/opt/cesga/anaconda/Anaconda2-5.3.0/lib/R'"
      ],
      "text/markdown": [
       "'/opt/cesga/anaconda/Anaconda2-5.3.0/lib/R'"
      ],
      "text/plain": [
       "[1] \"/opt/cesga/anaconda/Anaconda2-5.3.0/lib/R\""
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    ".libPaths()\n",
    "Sys.getenv(\"R_HOME\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### R session info"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "R version 3.5.1 (2018-07-02)\n",
       "Platform: x86_64-pc-linux-gnu (64-bit)\n",
       "Running under: CentOS Linux 7 (Core)\n",
       "\n",
       "Matrix products: default\n",
       "BLAS: /opt/cesga/anaconda/Anaconda2-5.3.0/lib/R/lib/libRblas.so\n",
       "LAPACK: /opt/cesga/anaconda/Anaconda2-5.3.0/lib/R/lib/libRlapack.so\n",
       "\n",
       "locale:\n",
       "[1] es_ES.UTF-8\n",
       "\n",
       "attached base packages:\n",
       "[1] stats     graphics  grDevices utils     datasets  methods   base     \n",
       "\n",
       "other attached packages:\n",
       "[1] bindrcpp_0.2.2       repr_0.15.0          ggplot2_3.0.0       \n",
       "[4] knitr_1.20           dplyr_0.7.6          sparklyr_0.8.4      \n",
       "[7] SparkR_1.6.1         RevoUtils_11.0.1     RevoUtilsMath_11.0.0\n",
       "\n",
       "loaded via a namespace (and not attached):\n",
       " [1] pbdZMQ_0.3-3     tidyselect_0.2.4 purrr_0.2.5      lattice_0.20-35 \n",
       " [5] colorspace_1.3-2 htmltools_0.3.6  yaml_2.2.0       base64enc_0.1-3 \n",
       " [9] rlang_0.2.1      later_0.7.3      pillar_1.3.0     withr_2.1.2     \n",
       "[13] glue_1.3.0       DBI_1.0.0        dbplyr_1.2.2     uuid_0.1-2      \n",
       "[17] bindr_0.1.1      plyr_1.8.4       stringr_1.3.1    munsell_0.5.0   \n",
       "[21] gtable_0.2.0     evaluate_0.11    labeling_0.3     httpuv_1.4.5    \n",
       "[25] parallel_3.5.1   broom_0.5.0      IRdisplay_0.5.0  Rcpp_0.12.18    \n",
       "[29] xtable_1.8-2     openssl_1.0.2    promises_1.0.1   backports_1.1.2 \n",
       "[33] scales_0.5.0     IRkernel_0.8.11  jsonlite_1.5     config_0.3      \n",
       "[37] mime_0.5         digest_0.6.15    stringi_1.2.4    shiny_1.1.0     \n",
       "[41] rprojroot_1.3-2  grid_3.5.1       tools_3.5.1      magrittr_1.5    \n",
       "[45] lazyeval_0.2.1   tibble_1.4.2     crayon_1.3.4     tidyr_0.8.1     \n",
       "[49] pkgconfig_2.0.1  assertthat_0.2.0 httr_1.3.1       rstudioapi_0.8  \n",
       "[53] R6_2.2.2         nlme_3.1-137     compiler_3.5.1  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sessionInfo()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## References \n",
    "\n",
    "#### Sparklyr\n",
    "\n",
    "- [`sparklyr` workshop](https://github.com/aurora-mareviv/sparklyr_start)\n",
    "- [`sparklyr` presentation by RStudio](https://cdn.oreillystatic.com/en/assets/1/event/193/Sparklyr_%20An%20R%20interface%20for%20Apache%20Spark%20Presentation.pdf)\n",
    "- [`sparklyr` tutorial](http://spark.rstudio.com/).\n",
    "- [`sparklyr` cheatsheet](https://www.rstudio.com/resources/cheatsheets/#sparklyr).\n",
    "- [`sparklyr`: creating extensions](http://spark.rstudio.com/extensions.html).\n",
    "- [Differences between `sparkr` and `sparklyr`](https://stackoverflow.com/questions/39494484/sparkr-vs-sparklyr).\n",
    "- [Hive Operators and UDFs](https://cwiki.apache.org/confluence/display/Hive/LanguageManual+UDF).\n",
    "- [String Functions in Hive](http://www.folkstalk.com/2011/11/string-functions-in-hive.html).\n",
    "- [POSIX regular expressions](https://www.postgresql.org/docs/9.4/static/functions-matching.html#FUNCTIONS-POSIX-REGEXP).\n",
    "\n",
    "#### PySpark\n",
    "\n",
    "- [PySpark course](https://github.com/javicacheiro/pyspark_course)\n",
    "- [Material for Machine Learning Workshop Galicia 2016](http://nbviewer.jupyter.org/github/javicacheiro/machine_learning_galicia_2016/blob/master/notebooks/sentiment_analysis-amazon_books.ipynb).\n",
    "- [PySpark Programming Guide](https://spark.apache.org/docs/0.9.0/python-programming-guide.html).\n",
    "- [PySpark cheatsheet](https://s3.amazonaws.com/assets.datacamp.com/blog_assets/PySpark_SQL_Cheat_Sheet_Python.pdf).\n",
    "\n",
    "#### Jupyter\n",
    "- [Jupyter shortcuts](https://www.dataquest.io/blog/jupyter-notebook-tips-tricks-shortcuts/)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Author: A. Baluja"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "R",
   "language": "R",
   "name": "ir"
  },
  "language_info": {
   "codemirror_mode": "r",
   "file_extension": ".r",
   "mimetype": "text/x-r-source",
   "name": "R",
   "pygments_lexer": "r",
   "version": "3.5.1"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}