clinUtils (version 0.1.4)

compareTables: Compare tables

Description

Compare tables

Usage

compareTables(
  newData,
  oldData,
  referenceVars = intersect(colnames(newData), colnames(oldData)),
  changeableVars = NULL,
  outputType = c("table-comparison", "newData-diff", "oldData-diff",
    "table-comparison-interactive", "newData-diff-interactive",
    "oldData-diff-interactive"),
  ...
)

Value

One of the output types specified in outputType. By default, all outputs are returned. If multiple output types are specified, a list of those are returned (named by output type).

Arguments

newData

data.frame object representing the new data

oldData

data.frame object representing the old data

referenceVars

character vector of the columns in the data that are the used as reference for the comparison.
If not specified, all columns present both in newData and oldData are considered.

changeableVars

character vector of the columns in the data for which you want to assess the change, e.g. variables that might have changed from the old to the new data.
If not specified, only 'Addition' and 'Removal' are detected.

outputType

String describing which output should be returned, (multiple are possible), either:

• 'table-comparison': data.frame containing difference between two datasets, see 'output' of compareDiff function.

• 'table-comparison-interactive': datatable object with differences between the two datasets, see 'output' of exportDiffData.

• 'newData-diff' or 'oldData-diff': data.frame with new/old data respectively, containing the information if each record differs in the old/new datasets respectively. See output of mergeDiffWithData.

• 'newData-diff-interactive' or 'oldData-diff-interactive': datatable with new/old data respectively, containing the information if each record differs in the old/new datasets respectively. See output of exportDiffData.

...

Any parameters passed to the exportDiffData function. These are only used if 'table-comparison-interactive' is specified in outputType.

Identification of the differences between datasets

To identify the differences between datasets, the following steps are followed:

1. removal of records identical between the old and new dataset (will be considered as 'Identical' later on)

2. records with a reference value present in the old dataset but not in the new dataset are considered 'Removal'

3. records with a reference value present in the new dataset but not in the old dataset are considered 'Addition'

4. records with reference value present both in the new and old dataset, after filtering of identical records and with difference in the changeable variables are considered 'Change'

Author

Laure Cougnaud, Michela Pasetto

Examples

## Example 1
# In this case the referenceVar 'a' is the same
# the comparison highlights only as change in the variables 'c' and 'd' 
newData <- data.frame(
	"a" = c(1, 2, 3, 4),
	"b" = c(5, 6, 7, 8),
	"c" = rep(1, 4),
	"d" = rep(2, 4)
)
oldData <- data.frame(
	"a" = c(1, 2, 3, 4),
	"b" = c(3, 4, 7, 8),
	"c" = rep(2, 4),
	"d" = rep(1, 4)
)
compareTables(
	newData = newData,
	oldData = oldData,
	referenceVars = "a",
	changeableVars = c("c", "d")
)

## Example 2
# In this case the referenceVar 'a' changes in the last two rows
# the comparison highlights as change the second and third rows in the variables 'c' and 'd'
# whereas the last rows are additions/removals with respect to the reference 'a'
newData <- data.frame(
		"a" = c(7, 1, 2, 3, 4),
		"b" = c(2, 1, 6, 7, 8),
		"c" = rep(1, 5),
		"d" = rep(2, 5)
)
oldData <- data.frame(
		"a" = c(7, 1, 2, 5, 6),
		"b" = c(2, 3, 4, 7, 8),
		"c" = c(1, rep(2, 4)),
		"d" = c(2, rep(1, 4))
)
compareTables(
		newData = newData,
		oldData = oldData,
		referenceVars = "a",
		changeableVars = c("c", "d")
)

## Example 3
# In this case the referenceVar 'a' is the same
# also the variable 'c' is the same and it's the only changeable var evaluated
newData <- data.frame(
		"a" = c(1, 2, 3, 4),
		"b" = c(5, 6, 7, 8),
		"c" = rep(1, 4),
		"d" = rep(2, 4)
)
oldData <- data.frame(
		"a" = c(1, 2, 3, 4),
		"b" = c(3, 4, 7, 8),
		"c" = rep(1, 4),
		"d" = rep(1, 4)
)
compareTables(
		newData = newData,
		oldData = oldData,
		referenceVars = "a",
		changeableVars = "c"
)

if (FALSE) # due to time constraint in CRAN

## In case only a specific output should be returned:
newData <- data.frame(
	"a" = c(7, 1, 2, 3, 4),
	"b" = c(2, 1, 6, 7, 8),
	"c" = rep(1, 5),
	"d" = rep(2, 5)
)
oldData <- data.frame(
	"a" = c(7, 1, 2, 5, 6),
	"b" = c(2, 3, 4, 7, 8),
	"c" = c(1, rep(2, 4)),
	"d" = c(2, rep(1, 4))
)

# get only the differences between datasets:

# as a data.frame
compareTables(newData = newData, oldData = oldData, 
	referenceVars = "a", changeableVars = c("c", "d"), 
	outputType = "table-comparison")

# as an interactive DataTable
compareTables(newData = newData, oldData = oldData, 
	referenceVars = "a", changeableVars = c("c", "d"), 
	outputType = "table-comparison-interactive"
)
# only the new data
compareTables(
    newData = newData, oldData = oldData, 
	referenceVars = "a", changeableVars = c("c", "d"), 
	outputType = "newData-diff"
)
# only the new data in interactive mode
compareTables(
    newData = newData, oldData = oldData, 
    referenceVars = "a", changeableVars = c("c", "d"), 
    outputType = "newData-diff-interactive"
)
# only the new data in static and interactive mode
compareTables(
    newData = newData, oldData = oldData, 
    referenceVars = "a", changeableVars = c("c", "d"), 
    outputType = c("newData-diff", "newData-diff-interactive")
)
# only the old data
compareTables(newData = newData, oldData = oldData, 
	referenceVars = "a", changeableVars = c("c", "d"), 
	outputType = "oldData-diff"
)
# only the old data in interactive mode
compareTables(
    newData = newData, oldData = oldData, 
    referenceVars = "a", changeableVars = c("c", "d"), 
    outputType = "oldData-diff-interactive"
)
# only the old data in static and interactive mode
compareTables(
    newData = newData, oldData = oldData, 
    referenceVars = "a", changeableVars = c("c", "d"), 
    outputType = c("oldData-diff", "oldData-diff-interactive")
)


## no changeable vars

newData <- data.frame(
	"a" = c(7, 1, 2, 3, 4),
	"b" = c(2, 1, 6, 7, 8),
	"c" = rep(1, 5),
	"d" = rep(2, 5)
)
oldData <- data.frame(
	"a" = c(7, 1, 2, 5, 6),
	"b" = c(2, 3, 4, 7, 8),
	"c" = c(1, rep(2, 4)),
	"d" = c(2, rep(1, 4))
)

compareTables(newData = newData, oldData = oldData, 
	referenceVars = "a"
)

## duplicated records

# in case there are multiple records for the same reference variables,
# identical records are flagged as 'Identity' and reported in the table
# reporting differences; and the different record are flagged as 'Change', 'Addition' or 'Removal'
newData <- data.frame(
	"a" = c(7, 7),
	"b" = c(1, 2),
	"c" = c(1, 2),
	"d" = c(2, 3)
)
oldData <- data.frame(
	"a" = c(7, 7, 7),
	"b" = c(3, 4, 5),
	"c" = c(1, 3, 5),
	"d" = c(2, 4, 6)
)
compareTables(
	newData = newData, oldData = oldData, 
	referenceVars = "a", changeableVars = c("c", "d"), 
)

## with labels in the interactive format, see ? getClinDT

if (FALSE) # due to time constraint in CRAN
	
newData <- data.frame(
	"a" = c(7, 1, 2, 3, 4),
	"b" = c(2, 1, 6, 7, 8),
	"c" = rep(1, 5),
	"d" = rep(2, 5)
)
oldData <- data.frame(
	"a" = c(7, 1, 2, 5, 6),
	"b" = c(2, 3, 4, 7, 8),
	"c" = c(1, rep(2, 4)),
	"d" = c(2, rep(1, 4))
)
compareTables(
	newData = newData,
	oldData = oldData,
	referenceVars = "a",
	changeableVars = c("c", "d"),
	# parameters passed to datatable
	colnames = c(
		"My reference variable" = "a", 
		"Changeable variable c" = "c",
		"Changeable variable d" = "d"
	)
)