dssSubsetByClass: Subset any number of vectors and/or dataframes by the levels...

In IulianD/dsSwissKnifeClient: DataSHIELD Tools and Utilities - client side

Description

Alternative to the datashield function ds.SubsetByClass

Usage

dssSubsetByClass(
  x,
  subsets = "subClasses",
  variables = NULL,
  keep.cols = NULL,
  async = TRUE,
  wait = TRUE,
  datasources = NULL
)

Arguments

x	a vector containing one or more names of vectors or dataframes present on the remote nodes
subsets	a character, name of the new list containing the subsets - default "subClasses"
variables	a vector containing one or more names of dataframe columns or standalone factors by which to subset
keep.cols	a vector containing the names of the columns or vectors to keep in the result (by default all are kept)
async	same as in datashield.assign
wait	same as in datashield.assign
datasources	same as in datashield.assign

Value

It doesn't return anything of value, it creates a list containing dataframes (as many as the combinations of categories of "variables") on each node

Examples

# open a local pseudo connection:
opals <-  dssCreatePseudoServers('test', servers = 1)
# load iris and create a vector and a factor:
datashield.aggregate(opals['local1'], as.symbol('partialData("iris", 41, 150)'))
play_env <- opals$local1$envir
assign('a', play_env$iris$Petal.Length, envir = play_env)
assign('b', factor(c(rep('category1',10), rep('category2',100))), envir = play_env)
#now play around:
dssSubsetByClass('iris', variables = 'iris$Species', datasources = opals)
str(play_env$subClasses)
dssSubsetByClass('iris', variables = c('iris$Species', 'b'), datasources = opals)
str(play_env$subClasses)
dssSubsetByClass(c('iris', 'a'), variables = c('iris$Species', 'b'), keep.cols = c('Sepal.Length', 'Sepal.Width', 'a'), datasources = opals)
str(play_env$subClasses)
dssSubsetByClass('a', variables = 'b', keep.cols = c('a'), datasources = opals)
str(play_env$subClasses)

IulianD/dsSwissKnifeClient documentation built on June 23, 2020, 4:38 p.m.