From f4bc01f1f33a93e6affe5c8a3e33ffbd92d03f38 Mon Sep 17 00:00:00 2001 From: Yu ISHIKAWA Date: Wed, 12 Aug 2015 18:33:27 -0700 Subject: [SPARK-9855] [SPARKR] Add expression functions into SparkR whose params are simple I added lots of expression functions for SparkR. This PR includes only functions whose params are only `(Column)` or `(Column, Column)`. And I think we need to improve how to test those functions. However, it would be better to work on another issue. ## Diff Summary - Add lots of functions in `functions.R` and their generic in `generic.R` - Add aliases for `ceiling` and `sign` - Move expression functions from `column.R` to `functions.R` - Modify `rdname` from `column` to `functions` I haven't supported `not` function, because the name has a collesion with `testthat` package. I didn't think of the way to define it. ## New Supported Functions ``` approxCountDistinct ascii base64 bin bitwiseNOT ceil (alias: ceiling) crc32 dayofmonth dayofyear explode factorial hex hour initcap isNaN last_day length log2 ltrim md5 minute month negate quarter reverse round rtrim second sha1 signum (alias: sign) size soundex to_date trim unbase64 unhex weekofyear year datediff levenshtein months_between nanvl pmod ``` ## JIRA [[SPARK-9855] Add expression functions into SparkR whose params are simple - ASF JIRA](https://issues.apache.org/jira/browse/SPARK-9855) Author: Yu ISHIKAWA Closes #8123 from yu-iskw/SPARK-9855. --- R/pkg/DESCRIPTION | 1 + R/pkg/R/column.R | 81 ----------------- R/pkg/R/functions.R | 123 ++++++++++++++++++++++++++ R/pkg/R/generics.R | 185 ++++++++++++++++++++++++++++++++++++--- R/pkg/inst/tests/test_sparkSQL.R | 21 +++-- 5 files changed, 309 insertions(+), 102 deletions(-) create mode 100644 R/pkg/R/functions.R (limited to 'R/pkg') diff --git a/R/pkg/DESCRIPTION b/R/pkg/DESCRIPTION index 4949d86d20..83e6489721 100644 --- a/R/pkg/DESCRIPTION +++ b/R/pkg/DESCRIPTION @@ -29,6 +29,7 @@ Collate: 'client.R' 'context.R' 'deserialize.R' + 'functions.R' 'mllib.R' 'serialize.R' 'sparkR.R' diff --git a/R/pkg/R/column.R b/R/pkg/R/column.R index eeaf9f193b..328f595d08 100644 --- a/R/pkg/R/column.R +++ b/R/pkg/R/column.R @@ -60,12 +60,6 @@ operators <- list( ) column_functions1 <- c("asc", "desc", "isNull", "isNotNull") column_functions2 <- c("like", "rlike", "startsWith", "endsWith", "getField", "getItem", "contains") -functions <- c("min", "max", "sum", "avg", "mean", "count", "abs", "sqrt", - "first", "last", "lower", "upper", "sumDistinct", - "acos", "asin", "atan", "cbrt", "ceiling", "cos", "cosh", "exp", - "expm1", "floor", "log", "log10", "log1p", "rint", "sign", - "sin", "sinh", "tan", "tanh", "toDegrees", "toRadians") -binary_mathfunctions <- c("atan2", "hypot") createOperator <- function(op) { setMethod(op, @@ -111,33 +105,6 @@ createColumnFunction2 <- function(name) { }) } -createStaticFunction <- function(name) { - setMethod(name, - signature(x = "Column"), - function(x) { - if (name == "ceiling") { - name <- "ceil" - } - if (name == "sign") { - name <- "signum" - } - jc <- callJStatic("org.apache.spark.sql.functions", name, x@jc) - column(jc) - }) -} - -createBinaryMathfunctions <- function(name) { - setMethod(name, - signature(y = "Column"), - function(y, x) { - if (class(x) == "Column") { - x <- x@jc - } - jc <- callJStatic("org.apache.spark.sql.functions", name, y@jc, x) - column(jc) - }) -} - createMethods <- function() { for (op in names(operators)) { createOperator(op) @@ -148,12 +115,6 @@ createMethods <- function() { for (name in column_functions2) { createColumnFunction2(name) } - for (x in functions) { - createStaticFunction(x) - } - for (name in binary_mathfunctions) { - createBinaryMathfunctions(name) - } } createMethods() @@ -242,45 +203,3 @@ setMethod("%in%", jc <- callJMethod(x@jc, "in", table) return(column(jc)) }) - -#' Approx Count Distinct -#' -#' @rdname column -#' @return the approximate number of distinct items in a group. -setMethod("approxCountDistinct", - signature(x = "Column"), - function(x, rsd = 0.95) { - jc <- callJStatic("org.apache.spark.sql.functions", "approxCountDistinct", x@jc, rsd) - column(jc) - }) - -#' Count Distinct -#' -#' @rdname column -#' @return the number of distinct items in a group. -setMethod("countDistinct", - signature(x = "Column"), - function(x, ...) { - jcol <- lapply(list(...), function (x) { - x@jc - }) - jc <- callJStatic("org.apache.spark.sql.functions", "countDistinct", x@jc, - listToSeq(jcol)) - column(jc) - }) - -#' @rdname column -#' @aliases countDistinct -setMethod("n_distinct", - signature(x = "Column"), - function(x, ...) { - countDistinct(x, ...) - }) - -#' @rdname column -#' @aliases count -setMethod("n", - signature(x = "Column"), - function(x) { - count(x) - }) diff --git a/R/pkg/R/functions.R b/R/pkg/R/functions.R new file mode 100644 index 0000000000..a15d2d5da5 --- /dev/null +++ b/R/pkg/R/functions.R @@ -0,0 +1,123 @@ +# +# Licensed to the Apache Software Foundation (ASF) under one or more +# contributor license agreements. See the NOTICE file distributed with +# this work for additional information regarding copyright ownership. +# The ASF licenses this file to You under the Apache License, Version 2.0 +# (the "License"); you may not use this file except in compliance with +# the License. You may obtain a copy of the License at +# +# http://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, software +# distributed under the License is distributed on an "AS IS" BASIS, +# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +# See the License for the specific language governing permissions and +# limitations under the License. +# + +#' @include generics.R column.R +NULL + +#' @title S4 expression functions for DataFrame column(s) +#' @description These are expression functions on DataFrame columns + +functions1 <- c( + "abs", "acos", "approxCountDistinct", "ascii", "asin", "atan", + "avg", "base64", "bin", "bitwiseNOT", "cbrt", "ceil", "cos", "cosh", "count", + "crc32", "dayofmonth", "dayofyear", "exp", "explode", "expm1", "factorial", + "first", "floor", "hex", "hour", "initcap", "isNaN", "last", "last_day", + "length", "log", "log10", "log1p", "log2", "lower", "ltrim", "max", "md5", + "mean", "min", "minute", "month", "negate", "quarter", "reverse", + "rint", "round", "rtrim", "second", "sha1", "signum", "sin", "sinh", "size", + "soundex", "sqrt", "sum", "sumDistinct", "tan", "tanh", "toDegrees", + "toRadians", "to_date", "trim", "unbase64", "unhex", "upper", "weekofyear", + "year") +functions2 <- c( + "atan2", "datediff", "hypot", "levenshtein", "months_between", "nanvl", "pmod") + +createFunction1 <- function(name) { + setMethod(name, + signature(x = "Column"), + function(x) { + jc <- callJStatic("org.apache.spark.sql.functions", name, x@jc) + column(jc) + }) +} + +createFunction2 <- function(name) { + setMethod(name, + signature(y = "Column"), + function(y, x) { + if (class(x) == "Column") { + x <- x@jc + } + jc <- callJStatic("org.apache.spark.sql.functions", name, y@jc, x) + column(jc) + }) +} + +createFunctions <- function() { + for (name in functions1) { + createFunction1(name) + } + for (name in functions2) { + createFunction2(name) + } +} + +createFunctions() + +#' Approx Count Distinct +#' +#' @rdname functions +#' @return the approximate number of distinct items in a group. +setMethod("approxCountDistinct", + signature(x = "Column"), + function(x, rsd = 0.95) { + jc <- callJStatic("org.apache.spark.sql.functions", "approxCountDistinct", x@jc, rsd) + column(jc) + }) + +#' Count Distinct +#' +#' @rdname functions +#' @return the number of distinct items in a group. +setMethod("countDistinct", + signature(x = "Column"), + function(x, ...) { + jcol <- lapply(list(...), function (x) { + x@jc + }) + jc <- callJStatic("org.apache.spark.sql.functions", "countDistinct", x@jc, + listToSeq(jcol)) + column(jc) + }) + +#' @rdname functions +#' @aliases ceil +setMethod("ceiling", + signature(x = "Column"), + function(x) { + ceil(x) + }) + +#' @rdname functions +#' @aliases signum +setMethod("sign", signature(x = "Column"), + function(x) { + signum(x) + }) + +#' @rdname functions +#' @aliases countDistinct +setMethod("n_distinct", signature(x = "Column"), + function(x, ...) { + countDistinct(x, ...) + }) + +#' @rdname functions +#' @aliases count +setMethod("n", signature(x = "Column"), + function(x) { + count(x) + }) diff --git a/R/pkg/R/generics.R b/R/pkg/R/generics.R index 379a78b1d8..f11e7fcb6a 100644 --- a/R/pkg/R/generics.R +++ b/R/pkg/R/generics.R @@ -575,10 +575,6 @@ setGeneric("approxCountDistinct", function(x, ...) { standardGeneric("approxCoun #' @export setGeneric("asc", function(x) { standardGeneric("asc") }) -#' @rdname column -#' @export -setGeneric("avg", function(x, ...) { standardGeneric("avg") }) - #' @rdname column #' @export setGeneric("between", function(x, bounds) { standardGeneric("between") }) @@ -587,13 +583,10 @@ setGeneric("between", function(x, bounds) { standardGeneric("between") }) #' @export setGeneric("cast", function(x, dataType) { standardGeneric("cast") }) -#' @rdname column -#' @export -setGeneric("cbrt", function(x) { standardGeneric("cbrt") }) - #' @rdname column #' @export setGeneric("contains", function(x, ...) { standardGeneric("contains") }) + #' @rdname column #' @export setGeneric("countDistinct", function(x, ...) { standardGeneric("countDistinct") }) @@ -658,22 +651,190 @@ setGeneric("rlike", function(x, ...) { standardGeneric("rlike") }) #' @export setGeneric("startsWith", function(x, ...) { standardGeneric("startsWith") }) -#' @rdname column + +###################### Expression Function Methods ########################## + +#' @rdname functions +#' @export +setGeneric("ascii", function(x) { standardGeneric("ascii") }) + +#' @rdname functions +#' @export +setGeneric("avg", function(x, ...) { standardGeneric("avg") }) + +#' @rdname functions +#' @export +setGeneric("base64", function(x) { standardGeneric("base64") }) + +#' @rdname functions +#' @export +setGeneric("bin", function(x) { standardGeneric("bin") }) + +#' @rdname functions +#' @export +setGeneric("bitwiseNOT", function(x) { standardGeneric("bitwiseNOT") }) + +#' @rdname functions +#' @export +setGeneric("cbrt", function(x) { standardGeneric("cbrt") }) + +#' @rdname functions +#' @export +setGeneric("ceil", function(x) { standardGeneric("ceil") }) + +#' @rdname functions +#' @export +setGeneric("crc32", function(x) { standardGeneric("crc32") }) + +#' @rdname functions +#' @export +setGeneric("datediff", function(y, x) { standardGeneric("datediff") }) + +#' @rdname functions +#' @export +setGeneric("dayofmonth", function(x) { standardGeneric("dayofmonth") }) + +#' @rdname functions +#' @export +setGeneric("dayofyear", function(x) { standardGeneric("dayofyear") }) + +#' @rdname functions +#' @export +setGeneric("explode", function(x) { standardGeneric("explode") }) + +#' @rdname functions +#' @export +setGeneric("hex", function(x) { standardGeneric("hex") }) + +#' @rdname functions +#' @export +setGeneric("hour", function(x) { standardGeneric("hour") }) + +#' @rdname functions +#' @export +setGeneric("initcap", function(x) { standardGeneric("initcap") }) + +#' @rdname functions +#' @export +setGeneric("isNaN", function(x) { standardGeneric("isNaN") }) + +#' @rdname functions +#' @export +setGeneric("last_day", function(x) { standardGeneric("last_day") }) + +#' @rdname functions +#' @export +setGeneric("levenshtein", function(y, x) { standardGeneric("levenshtein") }) + +#' @rdname functions +#' @export +setGeneric("lower", function(x) { standardGeneric("lower") }) + +#' @rdname functions +#' @export +setGeneric("ltrim", function(x) { standardGeneric("ltrim") }) + +#' @rdname functions +#' @export +setGeneric("md5", function(x) { standardGeneric("md5") }) + +#' @rdname functions +#' @export +setGeneric("minute", function(x) { standardGeneric("minute") }) + +#' @rdname functions +#' @export +setGeneric("month", function(x) { standardGeneric("month") }) + +#' @rdname functions +#' @export +setGeneric("months_between", function(y, x) { standardGeneric("months_between") }) + +#' @rdname functions +#' @export +setGeneric("nanvl", function(y, x) { standardGeneric("nanvl") }) + +#' @rdname functions +#' @export +setGeneric("negate", function(x) { standardGeneric("negate") }) + +#' @rdname functions +#' @export +setGeneric("pmod", function(y, x) { standardGeneric("pmod") }) + +#' @rdname functions +#' @export +setGeneric("quarter", function(x) { standardGeneric("quarter") }) + +#' @rdname functions +#' @export +setGeneric("reverse", function(x) { standardGeneric("reverse") }) + +#' @rdname functions +#' @export +setGeneric("rtrim", function(x) { standardGeneric("rtrim") }) + +#' @rdname functions +#' @export +setGeneric("second", function(x) { standardGeneric("second") }) + +#' @rdname functions +#' @export +setGeneric("sha1", function(x) { standardGeneric("sha1") }) + +#' @rdname functions +#' @export +setGeneric("signum", function(x) { standardGeneric("signum") }) + +#' @rdname functions +#' @export +setGeneric("size", function(x) { standardGeneric("size") }) + +#' @rdname functions +#' @export +setGeneric("soundex", function(x) { standardGeneric("soundex") }) + +#' @rdname functions #' @export setGeneric("sumDistinct", function(x) { standardGeneric("sumDistinct") }) -#' @rdname column +#' @rdname functions #' @export setGeneric("toDegrees", function(x) { standardGeneric("toDegrees") }) -#' @rdname column +#' @rdname functions #' @export setGeneric("toRadians", function(x) { standardGeneric("toRadians") }) -#' @rdname column +#' @rdname functions +#' @export +setGeneric("to_date", function(x) { standardGeneric("to_date") }) + +#' @rdname functions +#' @export +setGeneric("trim", function(x) { standardGeneric("trim") }) + +#' @rdname functions +#' @export +setGeneric("unbase64", function(x) { standardGeneric("unbase64") }) + +#' @rdname functions +#' @export +setGeneric("unhex", function(x) { standardGeneric("unhex") }) + +#' @rdname functions #' @export setGeneric("upper", function(x) { standardGeneric("upper") }) +#' @rdname functions +#' @export +setGeneric("weekofyear", function(x) { standardGeneric("weekofyear") }) + +#' @rdname functions +#' @export +setGeneric("year", function(x) { standardGeneric("year") }) + + #' @rdname glm #' @export setGeneric("glm") diff --git a/R/pkg/inst/tests/test_sparkSQL.R b/R/pkg/inst/tests/test_sparkSQL.R index 7377fc8f1c..e6d3b21ff8 100644 --- a/R/pkg/inst/tests/test_sparkSQL.R +++ b/R/pkg/inst/tests/test_sparkSQL.R @@ -640,15 +640,18 @@ test_that("column operators", { test_that("column functions", { c <- SparkR:::col("a") - c2 <- min(c) + max(c) + sum(c) + avg(c) + count(c) + abs(c) + sqrt(c) - c3 <- lower(c) + upper(c) + first(c) + last(c) - c4 <- approxCountDistinct(c) + countDistinct(c) + cast(c, "string") - c5 <- n(c) + n_distinct(c) - c5 <- acos(c) + asin(c) + atan(c) + cbrt(c) - c6 <- ceiling(c) + cos(c) + cosh(c) + exp(c) + expm1(c) - c7 <- floor(c) + log(c) + log10(c) + log1p(c) + rint(c) - c8 <- sign(c) + sin(c) + sinh(c) + tan(c) + tanh(c) - c9 <- toDegrees(c) + toRadians(c) + c1 <- abs(c) + acos(c) + approxCountDistinct(c) + ascii(c) + asin(c) + atan(c) + c2 <- avg(c) + base64(c) + bin(c) + bitwiseNOT(c) + cbrt(c) + ceil(c) + cos(c) + c3 <- cosh(c) + count(c) + crc32(c) + dayofmonth(c) + dayofyear(c) + exp(c) + c4 <- explode(c) + expm1(c) + factorial(c) + first(c) + floor(c) + hex(c) + c5 <- hour(c) + initcap(c) + isNaN(c) + last(c) + last_day(c) + length(c) + c6 <- log(c) + (c) + log1p(c) + log2(c) + lower(c) + ltrim(c) + max(c) + md5(c) + c7 <- mean(c) + min(c) + minute(c) + month(c) + negate(c) + quarter(c) + c8 <- reverse(c) + rint(c) + round(c) + rtrim(c) + second(c) + sha1(c) + c9 <- signum(c) + sin(c) + sinh(c) + size(c) + soundex(c) + sqrt(c) + sum(c) + c10 <- sumDistinct(c) + tan(c) + tanh(c) + toDegrees(c) + toRadians(c) + c11 <- to_date(c) + trim(c) + unbase64(c) + unhex(c) + upper(c) + weekofyear(c) + c12 <- year(c) df <- jsonFile(sqlContext, jsonPath) df2 <- select(df, between(df$age, c(20, 30)), between(df$age, c(10, 20))) -- cgit v1.2.3