randomSplit

Return a list of randomly split dataframes with the provided weights.

Usage

randomSplit(x, weights, seed)

# S4 method for class 'SparkDataFrame,numeric'
randomSplit(x, weights, seed)

Arguments

x

A SparkDataFrame

weights

A vector of weights for splits, will be normalized if they don't sum to 1

seed

A seed to use for random split

Note

randomSplit since 2.0.0

See also

Other SparkDataFrame functions: SparkDataFrame-class, agg(), alias(), arrange(), as.data.frame(), attach,SparkDataFrame-method, broadcast(), cache(), checkpoint(), coalesce(), collect(), colnames(), coltypes(), createOrReplaceTempView(), crossJoin(), cube(), dapply(), dapplyCollect(), describe(), dim(), distinct(), drop(), dropDuplicates(), dropna(), dtypes(), except(), exceptAll(), explain(), filter(), first(), gapply(), gapplyCollect(), getNumPartitions(), group_by(), head(), hint(), histogram(), insertInto(), intersect(), intersectAll(), isLocal(), isStreaming(), join(), limit(), localCheckpoint(), merge(), mutate(), ncol(), nrow(), persist(), printSchema(), rbind(), rename(), repartition(), repartitionByRange(), rollup(), sample(), saveAsTable(), schema(), select(), selectExpr(), show(), showDF(), storageLevel(), str(), subset(), summary(), take(), toJSON(), union(), unionAll(), unionByName(), unpersist(), unpivot(), with(), withColumn(), withWatermark(), write.df(), write.jdbc(), write.json(), write.orc(), write.parquet(), write.stream(), write.text()

Examples

if (FALSE) { # \dontrun{
sparkR.session()
df <- createDataFrame(data.frame(id = 1:1000))
df_list <- randomSplit(df, c(2, 3, 5), 0)
# df_list contains 3 SparkDataFrames with each having about 200, 300 and 500 rows respectively
sapply(df_list, count)
} # }