Major Description
Running build we have a bunch of warnings from using the `iris` dataset, for example.
Warning in FUN(X[[1L]], ...) :
Use Sepal_Length instead of Sepal.Length as column name
Warning in FUN(X[[2L]], ...) :
Use Sepal_Width instead of Sepal.Width as column name
Warning in FUN(X[[3L]], ...) :
Use Petal_Length instead of Petal.Length as column name
Warning in FUN(X[[4L]], ...) :
Use Petal_Width instead of Petal.Width as column name
Warning in FUN(X[[1L]], ...) :
Use Sepal_Length instead of Sepal.Length as column name
Warning in FUN(X[[2L]], ...) :
Use Sepal_Width instead of Sepal.Width as column name
Warning in FUN(X[[3L]], ...) :
Use Petal_Length instead of Petal.Length as column name
Warning in FUN(X[[4L]], ...) :
Use Petal_Width instead of Petal.Width as column name
Warning in FUN(X[[1L]], ...) :
Use Sepal_Length instead of Sepal.Length as column name
Warning in FUN(X[[2L]], ...) :
Use Sepal_Width instead of Sepal.Width as column name
Warning in FUN(X[[3L]], ...) :
Use Petal_Length instead of Petal.Length as column name
These are the results of having `.` in the column name. For reference, see SPARK-12191, SPARK-11976. Since it involves changing SQL, if we couldn't support that there then we should strongly consider using other dataset without `.`, eg. `cars`
And we should update this in API doc (roxygen2 doc string), vignettes, programming guide, R code example.
