NMdata (version 0.2.3)

mergeCoal: Overwrite values in one data.frame as available in another.

Description

Repair data.frame x with values present in data.frame y. Columns to match by must be provided.

Usage

mergeCoal(x, y, by, cols.coal, add.new = TRUE, as.fun)

Arguments

x

The initial data.frame

y

A data.frame to prioritize overc `x`.

by

Columns to merge by. A character vector, with names to columns present in both `x` and `y`. At least one of by and cols.merge must be provided.

cols.coal

Columns to overwrite values from `y` if available. cols.coal must be present in y and may be present in x.

add.new

If columns in y are ne to x, merge them in? Default is to do so.

as.fun

Pass a function (say tibble::as_tibble) in as.fun to convert to something else. If data.tables are wanted, use as.fun="data.table". The default is to return data as a data.frame. Modify the defaul using `NMdataConf()`.

Details

Non-na values in y will be used o overwrite columns in x at the rows matched using `by` columns.

Merges must be done using the same "by" columns for all rows. If rows needs to be merged using varying by columns, the merges must be done sequentially.

Will try to guess by and cols.coal.

Examples

library(data.table)
x <- data.table(idx=1:3,a=paste0("xa",1:3),b=paste0("xb",1:3))
y <- data.table(idx=1:2,a=c("ya1",NA),b=c(NA,"yb2"))
mergeCoal(x,y,by="idx")
## multiple rows in x matched by one row in y
x <- data.table(idx=1:3,grp=c(1,1,2),a=paste0("xa",1:3),b=paste0("xb",1:3))
y <- data.table(grp=1,a="y1")
mergeCoal(x,y,by="grp")
## new column (c in y, not in x)
x <- data.table(idx=1:3,a=paste0("xa",1:3),b=paste0("xb",1:3))
y <- data.table(idx=1:2,a=c("ya1",NA),b=c(NA,"yb2"),c=1)
mergeCoal(x,y,by="idx")