A custom indexer for the Datastep Array. The indexer will
return a value for all columns or a specified column. To
access all columns, leave the indexer empty. Otherwise, specify the
the column name(s) or number(s) to return data for. The indexer will
always act upon the current row in the datastep.
For additional details, see the dsarray function.
# S3 method for dsarray
[(x, i = NULL)The dsarray object.
The index of the datastep array item to return a value for. This index can be a column name or position in the array. It can also be a vector of column names or positions. If no index is supplied, a vector of all array values will be returned.
The value of the specified column for the current row in the datastep. If no index is supplied, a vector of all column values will be returned.
library(libr)
# Create AirPassengers Data Frame
df <- as.data.frame(t(matrix(AirPassengers, 12,
dimnames = list(month.abb, seq(1949, 1960)))),
stringsAsFactors = FALSE)
# Use datastep array to get sums by quarter
# Examine different ways of referencing data inside datastep
dat <- datastep(df,
keep = c("Q1", "Q2", "Q3", "Q4", "Tot"),
arrays = list(months = dsarray(names(df))),
{
# Reference by column name
Q1 <- Jan + Feb + Mar
# Reference by array positions
Q2 <- sum(months[4:6])
# Reference by array names
Q3 <- sum(months[c("Jul", "Aug", "Sep")])
# Reference by row position
Q4 <- rw$Oct + rw[["Nov"]] + rw[[12]]
# Empty indexer returns all column values in array
Tot <- sum(months[])
})
dat
# Q1 Q2 Q3 Q4 Tot
# 1949 362 385 432 341 1520
# 1950 382 409 498 387 1676
# 1951 473 513 582 474 2042
# 1952 544 582 681 557 2364
# 1953 628 707 773 592 2700
# 1954 627 725 854 661 2867
# 1955 742 854 1023 789 3408
# 1956 878 1005 1173 883 3939
# 1957 972 1125 1336 988 4421
# 1958 1020 1146 1400 1006 4572
# 1959 1108 1288 1570 1174 5140
# 1960 1227 1468 1736 1283 5714