[SPARK-11339][SPARKR] Document the list of functions in R base package that are masked by functions with same name in SparkR

Added tests for function that are reported as masked, to make sure the base:: or stats:: function can be called.

For those we can't call, added them to SparkR programming guide.

It would seem to me `table, sample, subset, filter, cov` not working are not actually expected - I investigated/experimented with them but couldn't get them to work. It looks like as they are defined in base or stats they are missing the S3 generic, eg.
```
> methods("transform")
[1] transform,ANY-method       transform.data.frame
[3] transform,DataFrame-method transform.default
see '?methods' for accessing help and source code
> methods("subset")
[1] subset.data.frame       subset,DataFrame-method subset.default
[4] subset.matrix
see '?methods' for accessing help and source code
Warning message:
In .S3methods(generic.function, class, parent.frame()) :
  function 'subset' appears not to be S3 generic; found functions that look like S3 methods
```
Any idea?

More information on masking:
http://www.ats.ucla.edu/stat/r/faq/referencing_objects.htm
http://www.sfu.ca/~sweldon/howTo/guide4.pdf

This is what the output doc looks like (minus css):
![image](https://cloud.githubusercontent.com/assets/8969467/11229714/2946e5de-8d4d-11e5-94b0-dda9696b6fdd.png)

Author: felixcheung <[email protected]>

Closes #9785 from felixcheung/rmasked.

Author: felixcheung

R/pkg/R/DataFrame.R

@@ -2152,7 +2152,7 @@ setMethod("with",
           })
 
 #' Returns the column types of a DataFrame.
-#' 
+#'
 #' @name coltypes
 #' @title Get column types of a DataFrame
 #' @family dataframe_funcs

R/pkg/R/functions.R

@@ -2204,7 +2204,7 @@ setMethod("denseRank",
 #' @export
 #' @examples \dontrun{lag(df$c)}
 setMethod("lag",
-          signature(x = "characterOrColumn", offset = "numeric", defaultValue = "ANY"),
+          signature(x = "characterOrColumn"),
           function(x, offset, defaultValue = NULL) {
             col <- if (class(x) == "Column") {
               x@jc

R/pkg/R/generics.R

@@ -539,7 +539,7 @@ setGeneric("showDF", function(x,...) { standardGeneric("showDF") })
 
 # @rdname subset
 # @export
-setGeneric("subset", function(x, subset, select, ...) { standardGeneric("subset") })
+setGeneric("subset", function(x, ...) { standardGeneric("subset") })
 
 #' @rdname agg
 #' @export
@@ -790,7 +790,7 @@ setGeneric("kurtosis", function(x) { standardGeneric("kurtosis") })
 
 #' @rdname lag
 #' @export
-setGeneric("lag", function(x, offset, defaultValue = NULL) { standardGeneric("lag") })
+setGeneric("lag", function(x, ...) { standardGeneric("lag") })
 
 #' @rdname last
 #' @export

R/pkg/inst/tests/test_mllib.R

@@ -31,6 +31,11 @@ test_that("glm and predict", {
   model <- glm(Sepal_Width ~ Sepal_Length, training, family = "gaussian")
   prediction <- predict(model, test)
   expect_equal(typeof(take(select(prediction, "prediction"), 1)$prediction), "double")
+
+  # Test stats::predict is working
+  x <- rnorm(15)
+  y <- x + rnorm(15)
+  expect_equal(length(predict(lm(y ~ x))), 15)
 })
 
 test_that("glm should work with long formula", {

R/pkg/inst/tests/test_sparkSQL.R

@@ -433,6 +433,10 @@ test_that("table() returns a new DataFrame", {
   expect_is(tabledf, "DataFrame")
   expect_equal(count(tabledf), 3)
   dropTempTable(sqlContext, "table1")
+
+  # Test base::table is working
+  #a <- letters[1:3]
+  #expect_equal(class(table(a, sample(a))), "table")
 })
 
 test_that("toRDD() returns an RRDD", {
@@ -673,6 +677,9 @@ test_that("sample on a DataFrame", {
   # Also test sample_frac
   sampled3 <- sample_frac(df, FALSE, 0.1, 0) # set seed for predictable result
   expect_true(count(sampled3) < 3)
+
+  # Test base::sample is working
+  #expect_equal(length(sample(1:12)), 12)
 })
 
 test_that("select operators", {
@@ -753,6 +760,9 @@ test_that("subsetting", {
   df6 <- subset(df, df$age %in% c(30), c(1,2))
   expect_equal(count(df6), 1)
   expect_equal(columns(df6), c("name", "age"))
+
+  # Test base::subset is working
+  expect_equal(nrow(subset(airquality, Temp > 80, select = c(Ozone, Temp))), 68)
 })
 
 test_that("selectExpr() on a DataFrame", {
@@ -888,6 +898,9 @@ test_that("column functions", {
   expect_equal(result, list(list(3L, 2L, 1L), list(6L, 5L, 4L)))
   result <- collect(select(df, sort_array(df[[1]])))[[1]]
   expect_equal(result, list(list(1L, 2L, 3L), list(4L, 5L, 6L)))
+
+  # Test that stats::lag is working
+  expect_equal(length(lag(ldeaths, 12)), 72)
 })
 #
 test_that("column binary mathfunctions", {
@@ -1086,7 +1099,7 @@ test_that("group by, agg functions", {
   gd3_local <- collect(agg(gd3, var(df8$age)))
   expect_equal(162, gd3_local[gd3_local$name == "Justin",][1, 2])
 
-  # make sure base:: or stats::sd, var are working
+  # Test stats::sd, stats::var are working
   expect_true(abs(sd(1:2) - 0.7071068) < 1e-6)
   expect_true(abs(var(1:5, 1:5) - 2.5) < 1e-6)
 
@@ -1138,6 +1151,9 @@ test_that("filter() on a DataFrame", {
   expect_equal(count(filtered5), 1)
   filtered6 <- where(df, df$age %in% c(19, 30))
   expect_equal(count(filtered6), 2)
+
+  # Test stats::filter is working
+  #expect_true(is.ts(filter(1:100, rep(1, 3))))
 })
 
 test_that("join() and merge() on a DataFrame", {
@@ -1284,6 +1300,12 @@ test_that("unionAll(), rbind(), except(), and intersect() on a DataFrame", {
   expect_is(unioned, "DataFrame")
   expect_equal(count(intersected), 1)
   expect_equal(first(intersected)$name, "Andy")
+
+  # Test base::rbind is working
+  expect_equal(length(rbind(1:4, c = 2, a = 10, 10, deparse.level = 0)), 16)
+
+  # Test base::intersect is working
+  expect_equal(length(intersect(1:20, 3:23)), 18)
 })
 
 test_that("withColumn() and withColumnRenamed()", {
@@ -1365,6 +1387,9 @@ test_that("describe() and summarize() on a DataFrame", {
   stats2 <- summary(df)
   expect_equal(collect(stats2)[4, "name"], "Andy")
   expect_equal(collect(stats2)[5, "age"], "30")
+
+  # Test base::summary is working
+  expect_equal(length(summary(attenu, digits = 4)), 35)
 })
 
 test_that("dropna() and na.omit() on a DataFrame", {
@@ -1448,6 +1473,9 @@ test_that("dropna() and na.omit() on a DataFrame", {
   expect_identical(expected, actual)
   actual <- collect(na.omit(df, minNonNulls = 3, cols = c("name", "age", "height")))
   expect_identical(expected, actual)
+
+  # Test stats::na.omit is working
+  expect_equal(nrow(na.omit(data.frame(x = c(0, 10, NA)))), 2)
 })
 
 test_that("fillna() on a DataFrame", {
@@ -1510,6 +1538,9 @@ test_that("cov() and corr() on a DataFrame", {
   expect_true(abs(result - 1.0) < 1e-12)
   result <- corr(df, "singles", "doubles", "pearson")
   expect_true(abs(result - 1.0) < 1e-12)
+
+  # Test stats::cov is working
+  #expect_true(abs(max(cov(swiss)) - 1739.295) < 1e-3)
 })
 
 test_that("freqItems() on a DataFrame", {

docs/sparkr.md

@@ -286,7 +286,7 @@ head(teenagers)
 
 # Machine Learning
 
-SparkR allows the fitting of generalized linear models over DataFrames using the [glm()](api/R/glm.html) function. Under the hood, SparkR uses MLlib to train a model of the specified family. Currently the gaussian and binomial families are supported. We support a subset of the available R formula operators for model fitting, including '~', '.', ':', '+', and '-'. 
+SparkR allows the fitting of generalized linear models over DataFrames using the [glm()](api/R/glm.html) function. Under the hood, SparkR uses MLlib to train a model of the specified family. Currently the gaussian and binomial families are supported. We support a subset of the available R formula operators for model fitting, including '~', '.', ':', '+', and '-'.
 
 The [summary()](api/R/summary.html) function gives the summary of a model produced by [glm()](api/R/glm.html).
 
@@ -351,3 +351,38 @@ summary(model)
 ##Sepal_Width    0.404655
 {% endhighlight %}
 </div>
+
+# R Function Name Conflicts
+
+When loading and attaching a new package in R, it is possible to have a name [conflict](https://stat.ethz.ch/R-manual/R-devel/library/base/html/library.html), where a
+function is masking another function.
+
+The following functions are masked by the SparkR package:
+
+<table class="table">
+  <tr><th>Masked function</th><th>How to Access</th></tr>
+  <tr>
+    <td><code>cov</code> in <code>package:stats</code></td>
+    <td><code><pre>stats::cov(x, y = NULL, use = "everything",
+           method = c("pearson", "kendall", "spearman"))</pre></code></td>
+  </tr>
+  <tr>
+    <td><code>filter</code> in <code>package:stats</code></td>
+    <td><code><pre>stats::filter(x, filter, method = c("convolution", "recursive"),
+              sides = 2, circular = FALSE, init)</pre></code></td>
+  </tr>
+  <tr>
+    <td><code>sample</code> in <code>package:base</code></td>
+    <td><code>base::sample(x, size, replace = FALSE, prob = NULL)</code></td>
+  </tr>
+  <tr>
+    <td><code>table</code> in <code>package:base</code></td>
+    <td><code><pre>base::table(...,
+            exclude = if (useNA == "no") c(NA, NaN),
+            useNA = c("no", "ifany", "always"),
+            dnn = list.names(...), deparse.level = 1)</pre></code></td>
+  </tr>
+</table>
+
+You can inspect the search path in R with [`search()`](https://stat.ethz.ch/R-manual/R-devel/library/base/html/search.html)
+