SparkR (R on Spark)

灰太狼 2022-07-13 13:10 189阅读 0赞

http://spark.apache.org/docs/latest/sparkr.html

# SparkR (R on Spark) #

*  [Overview][]
 *  [SparkDataFrame][]
    
     *  [Starting Up: SparkSession][Starting Up_ SparkSession]
     *  [Starting Up from RStudio][]
     *  [Creating SparkDataFrames][]
        
         *  [From local data frames][]
         *  [From Data Sources][]
         *  [From Hive tables][]
     *  [SparkDataFrame Operations][]
        
         *  [Selecting rows, columns][Selecting rows_ columns]
         *  [Grouping, Aggregation][Grouping_ Aggregation]
         *  [Operating on Columns][]
         *  [Applying User-Defined Function][]
            
             *  [Run a given function on a large dataset using `dapply` or `dapplyCollect`][Run a given function on a large dataset using_dapply_or_dapplyCollect]
                
                 *  [dapply][]
                 *  [dapplyCollect][]
             *  [Run a given function on a large dataset grouping by input column(s) and using `gapply` or `gapplyCollect`][Run a given function on a large dataset grouping by input column_s_ and using_gapply_or_gapplyCollect]
                
                 *  [gapply][]
                 *  [gapplyCollect][]
             *  [Data type mapping between R and Spark][]
             *  [Run local R functions distributed using `spark.lapply`][Run local R functions distributed using_spark.lapply]
                
                 *  [spark.lapply][]
     *  [Running SQL Queries from SparkR][]
 *  [Machine Learning][]
    
     *  [Algorithms][]
     *  [Model persistence][]
 *  [R Function Name Conflicts][]
 *  [Migration Guide][]
    
     *  [Upgrading From SparkR 1.5.x to 1.6.x][]
     *  [Upgrading From SparkR 1.6.x to 2.0][]
     *  [Upgrading to SparkR 2.1.0][]

# Overview #

SparkR is an R package that provides a light-weight frontend to use Apache Spark from R. In Spark 2.1.0, SparkR provides a distributed data frame implementation that supports operations like selection, filtering, aggregation etc. (similar to R data frames, [dplyr][]) but on large datasets. SparkR also supports distributed machine learning using MLlib.

# SparkDataFrame #

A SparkDataFrame is a distributed collection of data organized into named columns. It is conceptually equivalent to a table in a relational database or a data frame in R, but with richer optimizations under the hood. SparkDataFrames can be constructed from a wide array of sources such as: structured data files, tables in Hive, external databases, or existing local R data frames.

All of the examples on this page use sample data included in R or the Spark distribution and can be run using the `./bin/sparkR` shell.

## Starting Up: SparkSession ##

The entry point into SparkR is the `SparkSession` which connects your R program to a Spark cluster. You can create a `SparkSession` using `sparkR.session` and pass in options such as the application name, any spark packages depended on, etc. Further, you can also work with SparkDataFrames via `SparkSession`. If you are working from the `sparkR` shell, the `SparkSession` should already be created for you, and you would not need to call `sparkR.session`.

sparkR.session()

## Starting Up from RStudio ##

You can also start SparkR from RStudio. You can connect your R program to a Spark cluster from RStudio, R shell, Rscript or other R IDEs. To start, make sure SPARK\_HOME is set in environment (you can check [Sys.getenv][]), load the SparkR package, and call `sparkR.session` as below. It will check for the Spark installation, and, if not found, it will be downloaded and cached automatically. Alternatively, you can also run `install.spark` manually.

In addition to calling `sparkR.session`, you could also specify certain Spark driver properties. Normally these [Application properties][] and [Runtime Environment][] cannot be set programmatically, as the driver JVM process would have been started, in this case SparkR takes care of this for you. To set them, pass them as you would other configuration properties in the `sparkConfig` argument to `sparkR.session()`.

if (nchar(Sys.getenv("SPARK_HOME")) < 1) { 
      Sys.setenv(SPARK_HOME = "/home/spark")
    }
    library(SparkR, lib.loc = c(file.path(Sys.getenv("SPARK_HOME"), "R", "lib")))
    sparkR.session(master = "local[*]", sparkConfig = list(spark.driver.memory = "2g"))

The following Spark driver properties can be set in `sparkConfig` with `sparkR.session` from RStudio:

<table style="max-width:100%; background-color:transparent; border-collapse:collapse; border-spacing:0px; width:914px; margin-bottom:20px"> 
 <tbody> 
  <tr> 
   <th style="padding:8px; line-height:20px; vertical-align:top; border-top-width:1px; border-top-style:solid; border-top-color:rgb(221,221,221)"> Property Name</th> 
   <th style="padding:8px; line-height:20px; vertical-align:top; border-top-width:1px; border-top-style:solid; border-top-color:rgb(221,221,221)"> Property group</th> 
   <th style="padding:8px; line-height:20px; vertical-align:top; border-top-width:1px; border-top-style:solid; border-top-color:rgb(221,221,221)"> <code style="padding:0px; font-family:Menlo,'Lucida Console',monospace; color:rgb(68,68,68); border:none">spark-submit</code>&nbsp;equivalent</th> 
  </tr> 
  <tr> 
   <td style="padding:8px; line-height:20px; vertical-align:top; border-top-width:1px; border-top-style:solid; border-top-color:rgb(221,221,221)"> <code style="padding:0px; font-family:Menlo,'Lucida Console',monospace; color:rgb(68,68,68); border:none">spark.master</code></td> 
   <td style="padding:8px; line-height:20px; vertical-align:top; border-top-width:1px; border-top-style:solid; border-top-color:rgb(221,221,221)"> Application Properties</td> 
   <td style="padding:8px; line-height:20px; vertical-align:top; border-top-width:1px; border-top-style:solid; border-top-color:rgb(221,221,221)"> <code style="padding:0px; font-family:Menlo,'Lucida Console',monospace; color:rgb(68,68,68); border:none">--master</code></td> 
  </tr> 
  <tr> 
   <td style="padding:8px; line-height:20px; vertical-align:top; border-top-width:1px; border-top-style:solid; border-top-color:rgb(221,221,221)"> <code style="padding:0px; font-family:Menlo,'Lucida Console',monospace; color:rgb(68,68,68); border:none">spark.yarn.keytab</code></td> 
   <td style="padding:8px; line-height:20px; vertical-align:top; border-top-width:1px; border-top-style:solid; border-top-color:rgb(221,221,221)"> Application Properties</td> 
   <td style="padding:8px; line-height:20px; vertical-align:top; border-top-width:1px; border-top-style:solid; border-top-color:rgb(221,221,221)"> <code style="padding:0px; font-family:Menlo,'Lucida Console',monospace; color:rgb(68,68,68); border:none">--keytab</code></td> 
  </tr> 
  <tr> 
   <td style="padding:8px; line-height:20px; vertical-align:top; border-top-width:1px; border-top-style:solid; border-top-color:rgb(221,221,221)"> <code style="padding:0px; font-family:Menlo,'Lucida Console',monospace; color:rgb(68,68,68); border:none">spark.yarn.principal</code></td> 
   <td style="padding:8px; line-height:20px; vertical-align:top; border-top-width:1px; border-top-style:solid; border-top-color:rgb(221,221,221)"> Application Properties</td> 
   <td style="padding:8px; line-height:20px; vertical-align:top; border-top-width:1px; border-top-style:solid; border-top-color:rgb(221,221,221)"> <code style="padding:0px; font-family:Menlo,'Lucida Console',monospace; color:rgb(68,68,68); border:none">--principal</code></td> 
  </tr> 
  <tr> 
   <td style="padding:8px; line-height:20px; vertical-align:top; border-top-width:1px; border-top-style:solid; border-top-color:rgb(221,221,221)"> <code style="padding:0px; font-family:Menlo,'Lucida Console',monospace; color:rgb(68,68,68); border:none">spark.driver.memory</code></td> 
   <td style="padding:8px; line-height:20px; vertical-align:top; border-top-width:1px; border-top-style:solid; border-top-color:rgb(221,221,221)"> Application Properties</td> 
   <td style="padding:8px; line-height:20px; vertical-align:top; border-top-width:1px; border-top-style:solid; border-top-color:rgb(221,221,221)"> <code style="padding:0px; font-family:Menlo,'Lucida Console',monospace; color:rgb(68,68,68); border:none">--driver-memory</code></td> 
  </tr> 
  <tr> 
   <td style="padding:8px; line-height:20px; vertical-align:top; border-top-width:1px; border-top-style:solid; border-top-color:rgb(221,221,221)"> <code style="padding:0px; font-family:Menlo,'Lucida Console',monospace; color:rgb(68,68,68); border:none">spark.driver.extraClassPath</code></td> 
   <td style="padding:8px; line-height:20px; vertical-align:top; border-top-width:1px; border-top-style:solid; border-top-color:rgb(221,221,221)"> Runtime Environment</td> 
   <td style="padding:8px; line-height:20px; vertical-align:top; border-top-width:1px; border-top-style:solid; border-top-color:rgb(221,221,221)"> <code style="padding:0px; font-family:Menlo,'Lucida Console',monospace; color:rgb(68,68,68); border:none">--driver-class-path</code></td> 
  </tr> 
  <tr> 
   <td style="padding:8px; line-height:20px; vertical-align:top; border-top-width:1px; border-top-style:solid; border-top-color:rgb(221,221,221)"> <code style="padding:0px; font-family:Menlo,'Lucida Console',monospace; color:rgb(68,68,68); border:none">spark.driver.extraJavaOptions</code></td> 
   <td style="padding:8px; line-height:20px; vertical-align:top; border-top-width:1px; border-top-style:solid; border-top-color:rgb(221,221,221)"> Runtime Environment</td> 
   <td style="padding:8px; line-height:20px; vertical-align:top; border-top-width:1px; border-top-style:solid; border-top-color:rgb(221,221,221)"> <code style="padding:0px; font-family:Menlo,'Lucida Console',monospace; color:rgb(68,68,68); border:none">--driver-java-options</code></td> 
  </tr> 
  <tr> 
   <td style="padding:8px; line-height:20px; vertical-align:top; border-top-width:1px; border-top-style:solid; border-top-color:rgb(221,221,221)"> <code style="padding:0px; font-family:Menlo,'Lucida Console',monospace; color:rgb(68,68,68); border:none">spark.driver.extraLibraryPath</code></td> 
   <td style="padding:8px; line-height:20px; vertical-align:top; border-top-width:1px; border-top-style:solid; border-top-color:rgb(221,221,221)"> Runtime Environment</td> 
   <td style="padding:8px; line-height:20px; vertical-align:top; border-top-width:1px; border-top-style:solid; border-top-color:rgb(221,221,221)"> <code style="padding:0px; font-family:Menlo,'Lucida Console',monospace; color:rgb(68,68,68); border:none">--driver-library-path</code></td> 
  </tr> 
 </tbody> 
</table>

## Creating SparkDataFrames ##

With a `SparkSession`, applications can create `SparkDataFrame`s from a local R data frame, from a [Hive table][], or from other [data sources][].

### From local data frames ###

The simplest way to create a data frame is to convert a local R data frame into a SparkDataFrame. Specifically we can use `as.DataFrame` or `createDataFrame` and pass in the local R data frame to create a SparkDataFrame. As an example, the following creates a `SparkDataFrame` based using the `faithful` dataset from R.

df <- as.DataFrame(faithful)
    
    # Displays the first part of the SparkDataFrame
    head(df)
    ## eruptions waiting
    ##1 3.600 79
    ##2 1.800 54
    ##3 3.333 74

### From Data Sources ###

SparkR supports operating on a variety of data sources through the `SparkDataFrame` interface. This section describes the general methods for loading and saving data using Data Sources. You can check the Spark SQL programming guide for more [specific options][] that are available for the built-in data sources.

The general method for creating SparkDataFrames from data sources is `read.df`. This method takes in the path for the file to load and the type of data source, and the currently active SparkSession will be used automatically. SparkR supports reading JSON, CSV and Parquet files natively, and through packages available from sources like [Third Party Projects][], you can find data source connectors for popular file formats like Avro. These packages can either be added by specifying `--packages` with `spark-submit` or `sparkR` commands, or if initializing SparkSession with `sparkPackages` parameter when in an interactive R shell or from RStudio.

sparkR.session(sparkPackages = "com.databricks:spark-avro_2.11:3.0.0")

We can see how to use data sources using an example JSON input file. Note that the file that is used here is *not* a typical JSON file. Each line in the file must contain a separate, self-contained valid JSON object. For more information, please see [JSON Lines text format, also called newline-delimited JSON][JSON Lines text format_ also called newline-delimited JSON]. As a consequence, a regular multi-line JSON file will most often fail.

people <- read.df("./examples/src/main/resources/people.json", "json")
    head(people)
    ## age name
    ##1 NA Michael
    ##2 30 Andy
    ##3 19 Justin
    
    # SparkR automatically infers the schema from the JSON file
    printSchema(people)
    # root
    # |-- age: long (nullable = true)
    # |-- name: string (nullable = true)
    
    # Similarly, multiple files can be read with read.json
    people <- read.json(c("./examples/src/main/resources/people.json", "./examples/src/main/resources/people2.json"))

The data sources API natively supports CSV formatted input files. For more information please refer to SparkR [read.df][] API documentation.

df <- read.df(csvPath, "csv", header = "true", inferSchema = "true", na.strings = "NA")

The data sources API can also be used to save out SparkDataFrames into multiple file formats. For example we can save the SparkDataFrame from the previous example to a Parquet file using `write.df`.

write.df(people, path = "people.parquet", source = "parquet", mode = "overwrite")

### From Hive tables ###

You can also create SparkDataFrames from Hive tables. To do this we will need to create a SparkSession with Hive support which can access tables in the Hive MetaStore. Note that Spark should have been built with [Hive support][] and more details can be found in the [SQL programming guide][]. In SparkR, by default it will attempt to create a SparkSession with Hive support enabled (`enableHiveSupport = TRUE`).

sparkR.session()
    
    sql("CREATE TABLE IF NOT EXISTS src (key INT, value STRING)")
    sql("LOAD DATA LOCAL INPATH 'examples/src/main/resources/kv1.txt' INTO TABLE src")
    
    # Queries can be expressed in HiveQL.
    results <- sql("FROM src SELECT key, value")
    
    # results is now a SparkDataFrame
    head(results)
    ## key value
    ## 1 238 val_238
    ## 2 86 val_86
    ## 3 311 val_311

## SparkDataFrame Operations ##

SparkDataFrames support a number of functions to do structured data processing. Here we include some basic examples and a complete list can be found in the [API][] docs:

### Selecting rows, columns ###

# Create the SparkDataFrame
    df <- as.DataFrame(faithful)
    
    # Get basic information about the SparkDataFrame
    df
    ## SparkDataFrame[eruptions:double, waiting:double]
    
    # Select only the "eruptions" column
    head(select(df, df$eruptions))
    ## eruptions
    ##1 3.600
    ##2 1.800
    ##3 3.333
    
    # You can also pass in column name as strings
    head(select(df, "eruptions"))
    
    # Filter the SparkDataFrame to only retain rows with wait times shorter than 50 mins
    head(filter(df, df$waiting < 50))
    ## eruptions waiting
    ##1 1.750 47
    ##2 1.750 47
    ##3 1.867 48

### Grouping, Aggregation ###

SparkR data frames support a number of commonly used functions to aggregate data after grouping. For example we can compute a histogram of the `waiting` time in the `faithful` dataset as shown below

# We use the `n` operator to count the number of times each waiting time appears
    head(summarize(groupBy(df, df$waiting), count = n(df$waiting)))
    ## waiting count
    ##1 70 4
    ##2 67 1
    ##3 69 2
    
    # We can also sort the output from the aggregation to get the most common waiting times
    waiting_counts <- summarize(groupBy(df, df$waiting), count = n(df$waiting))
    head(arrange(waiting_counts, desc(waiting_counts$count)))
    ## waiting count
    ##1 78 15
    ##2 83 14
    ##3 81 13

### Operating on Columns ###

SparkR also provides a number of functions that can directly applied to columns for data processing and during aggregation. The example below shows the use of basic arithmetic functions.

# Convert waiting time from hours to seconds.
    # Note that we can assign this to a new column in the same SparkDataFrame
    df$waiting_secs <- df$waiting * 60
    head(df)
    ## eruptions waiting waiting_secs
    ##1 3.600 79 4740
    ##2 1.800 54 3240
    ##3 3.333 74 4440

### Applying User-Defined Function ###

In SparkR, we support several kinds of User-Defined Functions:

#### Run a given function on a large dataset using `dapply` or `dapplyCollect` ####

##### dapply #####

Apply a function to each partition of a `SparkDataFrame`. The function to be applied to each partition of the `SparkDataFrame` and should have only one parameter, to which a `data.frame` corresponds to each partition will be passed. The output of function should be a `data.frame`. Schema specifies the row format of the resulting a `SparkDataFrame`. It must match to [data types][Data type mapping between R and Spark] of returned value.

# Convert waiting time from hours to seconds.
    # Note that we can apply UDF to DataFrame.
    schema <- structType(structField("eruptions", "double"), structField("waiting", "double"),
                         structField("waiting_secs", "double"))
    df1 <- dapply(df, function(x) {  x <- cbind(x, x$waiting * 60) }, schema)
    head(collect(df1))
    ## eruptions waiting waiting_secs
    ##1 3.600 79 4740
    ##2 1.800 54 3240
    ##3 3.333 74 4440
    ##4 2.283 62 3720
    ##5 4.533 85 5100
    ##6 2.883 55 3300

##### dapplyCollect #####

Like `dapply`, apply a function to each partition of a `SparkDataFrame` and collect the result back. The output of function should be a `data.frame`. But, Schema is not required to be passed. Note that `dapplyCollect` can fail if the output of UDF run on all the partition cannot be pulled to the driver and fit in driver memory.

# Convert waiting time from hours to seconds.
    # Note that we can apply UDF to DataFrame and return a R's data.frame
    ldf <- dapplyCollect(
             df,
             function(x) { 
               x <- cbind(x, "waiting_secs" = x$waiting * 60)
             })
    head(ldf, 3)
    ## eruptions waiting waiting_secs
    ##1 3.600 79 4740
    ##2 1.800 54 3240
    ##3 3.333 74 4440

#### Run a given function on a large dataset grouping by input column(s) and using `gapply` or `gapplyCollect` ####

##### gapply #####

Apply a function to each group of a `SparkDataFrame`. The function is to be applied to each group of the `SparkDataFrame` and should have only two parameters: grouping key and R `data.frame` corresponding to that key. The groups are chosen from `SparkDataFrame`s column(s). The output of function should be a `data.frame`. Schema specifies the row format of the resulting `SparkDataFrame`. It must represent R function’s output schema on the basis of Spark [data types][Data type mapping between R and Spark]. The column names of the returned `data.frame` are set by user.

# Determine six waiting times with the largest eruption time in minutes.
    schema <- structType(structField("waiting", "double"), structField("max_eruption", "double"))
    result <- gapply(
        df,
        "waiting",
        function(key, x) { 
            y <- data.frame(key, max(x$eruptions))
        },
        schema)
    head(collect(arrange(result, "max_eruption", decreasing = TRUE)))
    
    ## waiting max_eruption
    ##1 64 5.100
    ##2 69 5.067
    ##3 71 5.033
    ##4 87 5.000
    ##5 63 4.933
    ##6 89 4.900

##### gapplyCollect #####

Like `gapply`, applies a function to each partition of a `SparkDataFrame` and collect the result back to R data.frame. The output of the function should be a `data.frame`. But, the schema is not required to be passed. Note that `gapplyCollect` can fail if the output of UDF run on all the partition cannot be pulled to the driver and fit in driver memory.

# Determine six waiting times with the largest eruption time in minutes.
    result <- gapplyCollect(
        df,
        "waiting",
        function(key, x) { 
            y <- data.frame(key, max(x$eruptions))
            colnames(y) <- c("waiting", "max_eruption")
            y
        })
    head(result[order(result$max_eruption, decreasing = TRUE), ])
    
    ## waiting max_eruption
    ##1 64 5.100
    ##2 69 5.067
    ##3 71 5.033
    ##4 87 5.000
    ##5 63 4.933
    ##6 89 4.900

#### Data type mapping between R and Spark ####

<table style="max-width:100%; border-collapse:collapse; border-spacing:0px; width:914px; margin-bottom:20px; color:rgb(29,31,34); font-family:'Helvetica Neue',Helvetica,Arial,sans-serif; font-size:14px; orphans:2; widows:2; background-color:transparent"> 
 <tbody> 
  <tr> 
   <th style="padding:8px; line-height:20px; vertical-align:top; border-top-width:1px; border-top-style:solid; border-top-color:rgb(221,221,221)"> R</th> 
   <th style="padding:8px; line-height:20px; vertical-align:top; border-top-width:1px; border-top-style:solid; border-top-color:rgb(221,221,221)"> Spark</th> 
  </tr> 
  <tr> 
   <td style="padding:8px; line-height:20px; vertical-align:top; border-top-width:1px; border-top-style:solid; border-top-color:rgb(221,221,221)"> byte</td> 
   <td style="padding:8px; line-height:20px; vertical-align:top; border-top-width:1px; border-top-style:solid; border-top-color:rgb(221,221,221)"> byte</td> 
  </tr> 
  <tr> 
   <td style="padding:8px; line-height:20px; vertical-align:top; border-top-width:1px; border-top-style:solid; border-top-color:rgb(221,221,221)"> integer</td> 
   <td style="padding:8px; line-height:20px; vertical-align:top; border-top-width:1px; border-top-style:solid; border-top-color:rgb(221,221,221)"> integer</td> 
  </tr> 
  <tr> 
   <td style="padding:8px; line-height:20px; vertical-align:top; border-top-width:1px; border-top-style:solid; border-top-color:rgb(221,221,221)"> float</td> 
   <td style="padding:8px; line-height:20px; vertical-align:top; border-top-width:1px; border-top-style:solid; border-top-color:rgb(221,221,221)"> float</td> 
  </tr> 
  <tr> 
   <td style="padding:8px; line-height:20px; vertical-align:top; border-top-width:1px; border-top-style:solid; border-top-color:rgb(221,221,221)"> double</td> 
   <td style="padding:8px; line-height:20px; vertical-align:top; border-top-width:1px; border-top-style:solid; border-top-color:rgb(221,221,221)"> double</td> 
  </tr> 
  <tr> 
   <td style="padding:8px; line-height:20px; vertical-align:top; border-top-width:1px; border-top-style:solid; border-top-color:rgb(221,221,221)"> numeric</td> 
   <td style="padding:8px; line-height:20px; vertical-align:top; border-top-width:1px; border-top-style:solid; border-top-color:rgb(221,221,221)"> double</td> 
  </tr> 
  <tr> 
   <td style="padding:8px; line-height:20px; vertical-align:top; border-top-width:1px; border-top-style:solid; border-top-color:rgb(221,221,221)"> character</td> 
   <td style="padding:8px; line-height:20px; vertical-align:top; border-top-width:1px; border-top-style:solid; border-top-color:rgb(221,221,221)"> string</td> 
  </tr> 
  <tr> 
   <td style="padding:8px; line-height:20px; vertical-align:top; border-top-width:1px; border-top-style:solid; border-top-color:rgb(221,221,221)"> string</td> 
   <td style="padding:8px; line-height:20px; vertical-align:top; border-top-width:1px; border-top-style:solid; border-top-color:rgb(221,221,221)"> string</td> 
  </tr> 
  <tr> 
   <td style="padding:8px; line-height:20px; vertical-align:top; border-top-width:1px; border-top-style:solid; border-top-color:rgb(221,221,221)"> binary</td> 
   <td style="padding:8px; line-height:20px; vertical-align:top; border-top-width:1px; border-top-style:solid; border-top-color:rgb(221,221,221)"> binary</td> 
  </tr> 
  <tr> 
   <td style="padding:8px; line-height:20px; vertical-align:top; border-top-width:1px; border-top-style:solid; border-top-color:rgb(221,221,221)"> raw</td> 
   <td style="padding:8px; line-height:20px; vertical-align:top; border-top-width:1px; border-top-style:solid; border-top-color:rgb(221,221,221)"> binary</td> 
  </tr> 
  <tr> 
   <td style="padding:8px; line-height:20px; vertical-align:top; border-top-width:1px; border-top-style:solid; border-top-color:rgb(221,221,221)"> logical</td> 
   <td style="padding:8px; line-height:20px; vertical-align:top; border-top-width:1px; border-top-style:solid; border-top-color:rgb(221,221,221)"> boolean</td> 
  </tr> 
  <tr> 
   <td style="padding:8px; line-height:20px; vertical-align:top; border-top-width:1px; border-top-style:solid; border-top-color:rgb(221,221,221)"> <a href="https://stat.ethz.ch/R-manual/R-devel/library/base/html/DateTimeClasses.html" style="color:rgb(0,136,204); text-decoration:none" rel="nofollow">POSIXct</a></td> 
   <td style="padding:8px; line-height:20px; vertical-align:top; border-top-width:1px; border-top-style:solid; border-top-color:rgb(221,221,221)"> timestamp</td> 
  </tr> 
  <tr> 
   <td style="padding:8px; line-height:20px; vertical-align:top; border-top-width:1px; border-top-style:solid; border-top-color:rgb(221,221,221)"> <a href="https://stat.ethz.ch/R-manual/R-devel/library/base/html/DateTimeClasses.html" style="color:rgb(0,136,204); text-decoration:none" rel="nofollow">POSIXlt</a></td> 
   <td style="padding:8px; line-height:20px; vertical-align:top; border-top-width:1px; border-top-style:solid; border-top-color:rgb(221,221,221)"> timestamp</td> 
  </tr> 
  <tr> 
   <td style="padding:8px; line-height:20px; vertical-align:top; border-top-width:1px; border-top-style:solid; border-top-color:rgb(221,221,221)"> <a href="https://stat.ethz.ch/R-manual/R-devel/library/base/html/Dates.html" style="color:rgb(0,136,204); text-decoration:none" rel="nofollow">Date</a></td> 
   <td style="padding:8px; line-height:20px; vertical-align:top; border-top-width:1px; border-top-style:solid; border-top-color:rgb(221,221,221)"> date</td> 
  </tr> 
  <tr> 
   <td style="padding:8px; line-height:20px; vertical-align:top; border-top-width:1px; border-top-style:solid; border-top-color:rgb(221,221,221)"> array</td> 
   <td style="padding:8px; line-height:20px; vertical-align:top; border-top-width:1px; border-top-style:solid; border-top-color:rgb(221,221,221)"> array</td> 
  </tr> 
  <tr> 
   <td style="padding:8px; line-height:20px; vertical-align:top; border-top-width:1px; border-top-style:solid; border-top-color:rgb(221,221,221)"> list</td> 
   <td style="padding:8px; line-height:20px; vertical-align:top; border-top-width:1px; border-top-style:solid; border-top-color:rgb(221,221,221)"> array</td> 
  </tr> 
  <tr> 
   <td style="padding:8px; line-height:20px; vertical-align:top; border-top-width:1px; border-top-style:solid; border-top-color:rgb(221,221,221)"> env</td> 
   <td style="padding:8px; line-height:20px; vertical-align:top; border-top-width:1px; border-top-style:solid; border-top-color:rgb(221,221,221)"> map</td> 
  </tr> 
 </tbody> 
</table>

#### Run local R functions distributed using `spark.lapply` ####

##### spark.lapply #####

Similar to `lapply` in native R, `spark.lapply` runs a function over a list of elements and distributes the computations with Spark. Applies a function in a manner that is similar to `doParallel` or `lapply` to elements of a list. The results of all the computations should fit in a single machine. If that is not the case they can do something like `df <- createDataFrame(list)` and then use `dapply`

# Perform distributed training of multiple models with spark.lapply. Here, we pass
    # a read-only list of arguments which specifies family the generalized linear model should be.
    families <- c("gaussian", "poisson")
    train <- function(family) { 
      model <- glm(Sepal.Length ~ Sepal.Width + Species, iris, family = family)
      summary(model)
    }
    # Return a list of model's summaries
    model.summaries <- spark.lapply(families, train)
    
    # Print the summary of each model
    print(model.summaries)

## Running SQL Queries from SparkR ##

A SparkDataFrame can also be registered as a temporary view in Spark SQL and that allows you to run SQL queries over its data. The `sql`function enables applications to run SQL queries programmatically and returns the result as a `SparkDataFrame`.

# Load a JSON file
    people <- read.df("./examples/src/main/resources/people.json", "json")
    
    # Register this SparkDataFrame as a temporary view.
    createOrReplaceTempView(people, "people")
    
    # SQL statements can be run by using the sql method
    teenagers <- sql("SELECT name FROM people WHERE age >= 13 AND age <= 19")
    head(teenagers)
    ## name
    ##1 Justin

# Machine Learning #

## Algorithms ##

SparkR supports the following machine learning algorithms currently:

*  [`spark.glm`][spark.glm] or [`glm`][glm]: [`Generalized Linear Model`][Generalized Linear Model]
 *  [`spark.survreg`][spark.survreg]: [`Accelerated Failure Time (AFT) Survival Regression Model`][Accelerated Failure Time _AFT_ Survival Regression Model]
 *  [`spark.naiveBayes`][spark.naiveBayes]: [`Naive Bayes Model`][Naive Bayes Model]
 *  [`spark.kmeans`][spark.kmeans]: [`K-Means Model`][K-Means Model]
 *  [`spark.logit`][spark.logit]: [`Logistic Regression Model`][Logistic Regression Model]
 *  [`spark.isoreg`][spark.isoreg]: [`Isotonic Regression Model`][Isotonic Regression Model]
 *  [`spark.gaussianMixture`][spark.gaussianMixture]: [`Gaussian Mixture Model`][Gaussian Mixture Model]
 *  [`spark.lda`][spark.lda]: [`Latent Dirichlet Allocation (LDA) Model`][Latent Dirichlet Allocation _LDA_ Model]
 *  [`spark.mlp`][spark.mlp]: [`Multilayer Perceptron Classification Model`][Multilayer Perceptron Classification Model]
 *  [`spark.gbt`][spark.gbt]: `Gradient Boosted Tree Model for` [`Regression`][Regression] `and` [`Classification`][Classification]
 *  [`spark.randomForest`][spark.randomForest]: `Random Forest Model for` [`Regression`][Regression 1] `and` [`Classification`][Classification 1]
 *  [`spark.als`][spark.als]: [`Alternating Least Squares (ALS) matrix factorization Model`][Alternating Least Squares _ALS_ matrix factorization Model]
 *  [`spark.kstest`][spark.kstest]: `Kolmogorov-Smirnov Test`

Under the hood, SparkR uses MLlib to train the model. Please refer to the corresponding section of MLlib user guide for example code. Users can call `summary` to print a summary of the fitted model, [predict][] to make predictions on new data, and [write.ml][]/[read.ml][] to save/load fitted models. SparkR supports a subset of the available R formula operators for model fitting, including ‘~’, ‘.’, ‘:’, ‘+’, and ‘-‘.

## Model persistence ##

The following example shows how to save/load a MLlib model by SparkR. <div class="highlight"><pre>irisDF <-suppressWarnings(createDataFrame(iris)) *\# Fit a generalized linear model of family "gaussian" with spark.glm* gaussianDF <- irisDF gaussianTestDF <- irisDF gaussianGLM <- spark.glm(gaussianDF, Sepal\_Length ~ Sepal\_Width \+ Species, family = "gaussian")

*\# Save and then load a fitted MLlib model* modelPath <- tempfile(pattern = "ml", fileext = ".tmp") write.ml(gaussianGLM, modelPath)gaussianGLM2 <- read.ml(modelPath)

*\# Check model summary* summary(gaussianGLM2)

*\# Check model prediction* gaussianPredictions <- predict(gaussianGLM2, gaussianTestDF) showDF(gaussianPredictions)

unlink(modelPath) </pre></div><div>Find full example code at “examples/src/main/r/ml/ml.R” in the Spark repo.</div>

# R Function Name Conflicts #

When loading and attaching a new package in R, it is possible to have a name [conflict][], where a function is masking another function.

The following functions are masked by the SparkR package:

Since part of SparkR is modeled on the `dplyr` package, certain functions in SparkR share the same names with those in `dplyr`. Depending on the load order of the two packages, some functions from the package loaded first are masked by those in the package loaded after. In such case, prefix such calls with the package name, for instance, `SparkR::cume_dist(x)` or `dplyr::cume_dist(x)`.

You can inspect the search path in R with [`search()`][search]

# Migration Guide #

## Upgrading From SparkR 1.5.x to 1.6.x ##

*  Before Spark 1.6.0, the default mode for writes was `append`. It was changed in Spark 1.6.0 to `error` to match the Scala API.
 *  SparkSQL converts `NA` in R to `null` and vice-versa.

## Upgrading From SparkR 1.6.x to 2.0 ##

*  The method `table` has been removed and replaced by `tableToDF`.
 *  The class `DataFrame` has been renamed to `SparkDataFrame` to avoid name conflicts.
 *  Spark’s `SQLContext` and `HiveContext` have been deprecated to be replaced by `SparkSession`. Instead of `sparkR.init()`, call `sparkR.session()` in its place to instantiate the SparkSession. Once that is done, that currently active SparkSession will be used for SparkDataFrame operations.
 *  The parameter `sparkExecutorEnv` is not supported by `sparkR.session`. To set environment for the executors, set Spark config properties with the prefix “spark.executorEnv.VAR\_NAME”, for example, “spark.executorEnv.PATH”
 *  The `sqlContext` parameter is no longer required for these functions: `createDataFrame`, `as.DataFrame`, `read.json`, `jsonFile`, `read.parquet`, `parquetFile`, `read.text`, `sql`, `tables`, `tableNames`, `cacheTable`, `uncacheTable`, `clearCache`, `dropTempTable`, `read.df`, `loadDF`, `createExternalTable`.
 *  The method `registerTempTable` has been deprecated to be replaced by `createOrReplaceTempView`.
 *  The method `dropTempTable` has been deprecated to be replaced by `dropTempView`.
 *  The `sc` SparkContext parameter is no longer required for these functions: `setJobGroup`, `clearJobGroup`, `cancelJobGroup`

## Upgrading to SparkR 2.1.0 ##

*  `join` no longer performs Cartesian Product by default, use `crossJoin` instead.

[Overview]: http://spark.apache.org/docs/latest/sparkr.html#overview
[SparkDataFrame]: http://spark.apache.org/docs/latest/sparkr.html#sparkdataframe
[Starting Up_ SparkSession]: http://spark.apache.org/docs/latest/sparkr.html#starting-up-sparksession
[Starting Up from RStudio]: http://spark.apache.org/docs/latest/sparkr.html#starting-up-from-rstudio
[Creating SparkDataFrames]: http://spark.apache.org/docs/latest/sparkr.html#creating-sparkdataframes
[From local data frames]: http://spark.apache.org/docs/latest/sparkr.html#from-local-data-frames
[From Data Sources]: http://spark.apache.org/docs/latest/sparkr.html#from-data-sources
[From Hive tables]: http://spark.apache.org/docs/latest/sparkr.html#from-hive-tables
[SparkDataFrame Operations]: http://spark.apache.org/docs/latest/sparkr.html#sparkdataframe-operations
[Selecting rows_ columns]: http://spark.apache.org/docs/latest/sparkr.html#selecting-rows-columns
[Grouping_ Aggregation]: http://spark.apache.org/docs/latest/sparkr.html#grouping-aggregation
[Operating on Columns]: http://spark.apache.org/docs/latest/sparkr.html#operating-on-columns
[Applying User-Defined Function]: http://spark.apache.org/docs/latest/sparkr.html#applying-user-defined-function
[Run a given function on a large dataset using_dapply_or_dapplyCollect]: http://spark.apache.org/docs/latest/sparkr.html#run-a-given-function-on-a-large-dataset-using-dapply-or-dapplycollect
[dapply]: http://spark.apache.org/docs/latest/sparkr.html#dapply
[dapplyCollect]: http://spark.apache.org/docs/latest/sparkr.html#dapplycollect
[Run a given function on a large dataset grouping by input column_s_ and using_gapply_or_gapplyCollect]: http://spark.apache.org/docs/latest/sparkr.html#run-a-given-function-on-a-large-dataset-grouping-by-input-columns-and-using-gapply-or-gapplycollect
[gapply]: http://spark.apache.org/docs/latest/sparkr.html#gapply
[gapplyCollect]: http://spark.apache.org/docs/latest/sparkr.html#gapplycollect
[Data type mapping between R and Spark]: http://spark.apache.org/docs/latest/sparkr.html#data-type-mapping-between-r-and-spark
[Run local R functions distributed using_spark.lapply]: http://spark.apache.org/docs/latest/sparkr.html#run-local-r-functions-distributed-using-sparklapply
[spark.lapply]: http://spark.apache.org/docs/latest/sparkr.html#sparklapply
[Running SQL Queries from SparkR]: http://spark.apache.org/docs/latest/sparkr.html#running-sql-queries-from-sparkr
[Machine Learning]: http://spark.apache.org/docs/latest/sparkr.html#machine-learning
[Algorithms]: http://spark.apache.org/docs/latest/sparkr.html#algorithms
[Model persistence]: http://spark.apache.org/docs/latest/sparkr.html#model-persistence
[R Function Name Conflicts]: http://spark.apache.org/docs/latest/sparkr.html#r-function-name-conflicts
[Migration Guide]: http://spark.apache.org/docs/latest/sparkr.html#migration-guide
[Upgrading From SparkR 1.5.x to 1.6.x]: http://spark.apache.org/docs/latest/sparkr.html#upgrading-from-sparkr-15x-to-16x
[Upgrading From SparkR 1.6.x to 2.0]: http://spark.apache.org/docs/latest/sparkr.html#upgrading-from-sparkr-16x-to-20
[Upgrading to SparkR 2.1.0]: http://spark.apache.org/docs/latest/sparkr.html#upgrading-to-sparkr-210
[dplyr]: https://github.com/hadley/dplyr
[Sys.getenv]: https://stat.ethz.ch/R-manual/R-devel/library/base/html/Sys.getenv.html
[Application properties]: http://spark.apache.org/docs/latest/configuration.html#application-properties
[Runtime Environment]: http://spark.apache.org/docs/latest/configuration.html#runtime-environment
[Hive table]: http://spark.apache.org/docs/latest/sql-programming-guide.html#hive-tables
[data sources]: http://spark.apache.org/docs/latest/sql-programming-guide.html#data-sources
[specific options]: http://spark.apache.org/docs/latest/sql-programming-guide.html#manually-specifying-options
[Third Party Projects]: http://spark.apache.org/third-party-projects.html
[JSON Lines text format_ also called newline-delimited JSON]: http://jsonlines.org/
[read.df]: http://spark.apache.org/docs/latest/api/R/read.df.html
[Hive support]: http://spark.apache.org/docs/latest/building-spark.html#building-with-hive-and-jdbc-support
[SQL programming guide]: http://spark.apache.org/docs/latest/sql-programming-guide.html#starting-point-sparksession
[API]: http://spark.apache.org/docs/latest/api/R/index.html
[spark.glm]: http://spark.apache.org/docs/latest/api/R/spark.glm.html
[glm]: http://spark.apache.org/docs/latest/api/R/glm.html
[Generalized Linear Model]: http://spark.apache.org/docs/latest/ml-classification-regression.html#generalized-linear-regression
[spark.survreg]: http://spark.apache.org/docs/latest/api/R/spark.survreg.html
[Accelerated Failure Time _AFT_ Survival Regression Model]: http://spark.apache.org/docs/latest/ml-classification-regression.html#survival-regression
[spark.naiveBayes]: http://spark.apache.org/docs/latest/api/R/spark.naiveBayes.html
[Naive Bayes Model]: http://spark.apache.org/docs/latest/ml-classification-regression.html#naive-bayes
[spark.kmeans]: http://spark.apache.org/docs/latest/api/R/spark.kmeans.html
[K-Means Model]: http://spark.apache.org/docs/latest/ml-clustering.html#k-means
[spark.logit]: http://spark.apache.org/docs/latest/api/R/spark.logit.html
[Logistic Regression Model]: http://spark.apache.org/docs/latest/ml-classification-regression.html#logistic-regression
[spark.isoreg]: http://spark.apache.org/docs/latest/api/R/spark.isoreg.html
[Isotonic Regression Model]: http://spark.apache.org/docs/latest/ml-classification-regression.html#isotonic-regression
[spark.gaussianMixture]: http://spark.apache.org/docs/latest/api/R/spark.gaussianMixture.html
[Gaussian Mixture Model]: http://spark.apache.org/docs/latest/ml-clustering.html#gaussian-mixture-model-gmm
[spark.lda]: http://spark.apache.org/docs/latest/api/R/spark.lda.html
[Latent Dirichlet Allocation _LDA_ Model]: http://spark.apache.org/docs/latest/ml-clustering.html#latent-dirichlet-allocation-lda
[spark.mlp]: http://spark.apache.org/docs/latest/api/R/spark.mlp.html
[Multilayer Perceptron Classification Model]: http://spark.apache.org/docs/latest/ml-classification-regression.html#multilayer-perceptron-classifier
[spark.gbt]: http://spark.apache.org/docs/latest/api/R/spark.gbt.html
[Regression]: http://spark.apache.org/docs/latest/ml-classification-regression.html#gradient-boosted-tree-regression
[Classification]: http://spark.apache.org/docs/latest/ml-classification-regression.html#gradient-boosted-tree-classifier
[spark.randomForest]: http://spark.apache.org/docs/latest/api/R/spark.randomForest.html
[Regression 1]: http://spark.apache.org/docs/latest/ml-classification-regression.html#random-forest-regression
[Classification 1]: http://spark.apache.org/docs/latest/ml-classification-regression.html#random-forest-classifier
[spark.als]: http://spark.apache.org/docs/latest/api/R/spark.als.html
[Alternating Least Squares _ALS_ matrix factorization Model]: http://spark.apache.org/docs/latest/ml-collaborative-filtering.html#collaborative-filtering
[spark.kstest]: http://spark.apache.org/docs/latest/api/R/spark.kstest.html
[predict]: http://spark.apache.org/docs/latest/api/R/predict.html
[write.ml]: http://spark.apache.org/docs/latest/api/R/write.ml.html
[read.ml]: http://spark.apache.org/docs/latest/api/R/read.ml.html
[conflict]: https://stat.ethz.ch/R-manual/R-devel/library/base/html/library.html
[search]: https://stat.ethz.ch/R-manual/R-devel/library/base/html/search.html