defCasTable: Create a CASTable Object for an Existing CAS Table

In sassoftware/R-swat: SAS Scripting Wrapper for Analytics Transfer (SWAT)

Create a CASTable Object for an Existing CAS Table

Description

Creates a CASTable object to reference an existing in-memory table in CAS. You can use this function to reference tables that were loaded by other SAS products, other scripts, or from server-side loads with the cas.table.loadTable function.

Usage

defCasTable(
  conn,
  tablename,
  caslib = "",
  columns = "",
  where = "",
  orderby = list(),
  groupby = list(),
  gbmode = ""
)

Arguments

conn	A CAS object that represents a connection and session in CAS.
tablename	A character that specifies the in-memory table name. You can run the cas.table.tableInfo function to list the in-memory tables.
caslib	An optional character string that identifies the caslib for the in-memory table. Specify this parameter to override the active caslib.
columns	A list of column names.
where	A character string that specifies a filter for the rows to process. The filter uses syntax that is specific to SAS.
orderby	A list of column names. Rows are partitioned according to the columns in the groupby parameter and then ordered according to the values of the columns specified in this parameter.
groupby	A list of column names. If you specify this parameter when you load an in-memory table, then the table is partitioned by the columns. If you specify this parameter when running an action, then BY-groups are formed temporarily for the duration of the action.
gbmode	A character string. Values are NOSORT (default) or REDISTRIBUTE. See the CAS product documentation for more information.

Value

CASTable

Examples

## Not run: 
irisct <- as.casTable(s, iris, casOut="irisct")

# Create another CASTable instance to the same in-memory table,
# but specify that CAS actions are performed by groups of species.
irisct.grouped <- defCasTable(s, tablename="irisct", groupby=list("species"))

## End(Not run)

sassoftware/R-swat documentation built on Feb. 19, 2025, 10:33 a.m.