cross: Cross-tabulation
In dkneis/mcu: Model calibration utilities

Description Usage Arguments Value Author(s) Examples

Converts a data frame in long format (e.g. output from a data base) into wide format (cross-table) based on criteria from a single column.

1 2	cross(x, colsRowHead, colColHead, colValues, prefix = "", suffix = "", sortRows = 0, sortCols = 0, sep = "\t")

`x`	Data frame to be transformed.
`colsRowHead`	Columns in `x` forming the row headers of the result. These columns keep their position. They usually contain factor data such as strings, integers, or logical values.
`colColHead`	Single column in `x` defining the newly created column headers in the result. This column usually contains factor data such as strings, integers, or logical values.
`colValues`	Column with the actual data used to populate the resulting cross table.
`prefix`	Character string used as a prefix when constructing the names of the newly created columns. This is usually required to get valid column names if the data in column `colColHead` are not strings.
`suffix`	Character string used as a suffix when constructing the names of the newly created columns.
`sortRows`	Numeric. If zero, no sorting is performed. Values greater (less than) zero sort the result in ascending (descending) order based on the combined columns in `colsRowHead`.
`sortCols`	Numeric. If zero, no sorting is performed. Values greater (less than) zero sort the newly created columns in ascending (descending) order based on their names.
`sep`	A character string used as a separator when collapsing information from different columns. This string must not appear in any of the involved columns. An error is generated if this is the case.

A data frame with m + n columns where m is the length of colsRowHead and n equals the number of unique values in the column specified as colColHead.

david.kneis@tu-dresden.de

# Data base format: Meteorological variables at different times and locations
times= ISOdate(2000,6,1:10)
locs= c("Berlin","London","Rome")
vars= c("rain", "temp", "wind")
d= data.frame(stringsAsFactors=FALSE,
  time= rep(times, length(locs)*length(vars)),
  loc= rep(rep(locs, length(vars)), each=length(times)),
  var= rep(vars, each=length(times)*length(vars)),
  val= NA
)
d$val[d$var == "rain"]= runif(sum(d$var == "rain")) > 0.5
d$val[d$var == "temp"]= rnorm(n=sum(d$var == "temp"), mean=20, sd=5)
d$val[d$var == "wind"]= 10^runif(sum(d$var == "wind"), min=-3, max=2)

# Cross table with observed variables in columns
cross(x=d, colsRowHead=c("time","loc"), colColHead="var", colValues="val",
  prefix="", suffix="", sortRows=0, sortCols=0, sep="\t")

# Cross table with locations in columns
cross(x=d, colsRowHead=c("time","var"), colColHead="loc", colValues="val",
  prefix="", suffix="", sortRows=0, sortCols=0, sep="\t")

# A common mistake
## Not run: 
  cross(x=d, colsRowHead=c("time"), colColHead="loc", colValues="val",
    prefix="", suffix="", sortRows=0, sortCols=0, sep="\t")

## End(Not run)