audited-package: Automatic Tracing and Display of Changes to Row Count

In audited: Automatic Tracing and Display of Changes to Row Count

Description

audited classifies a data.frame such that row deletions and additions are tracked. A mechanism exists to give formal names to the row subsets added or deleted. These names are used to populate a directed graph giving an account of all the transactions contributing to the state of the object.

Details

Package:	audited
Type:	Package
Version:	1.10
Date:	2014-09-24
License:	GPL3

Audited data frames keep track of what happens to them, so later you can create a graph showing row deletions and additions.

as.audited creates an audited data frame (which is also keyed). You can supply a name (id) or accept the default (same as object name). When you add or delete rows using supported methods, a transaction will be stored in the audit table (bound to the object) indicating the number of changed rows and the resulting total. If the data frame has a column listed in the vector options('audit') (set this using options(audit=)), unique values of that column (first match) will be tracked also. If options('artifact') is TRUE (default NULL) then dropped record sets will be stored in a list as the artifact attribute of the audited data frame. (Alternatively, this can be a vector with any of drop, add, create, modify, transform or merge: see artifact for details.) as.xlsx writes artifacts to a workbook.

Author(s)

Tim Bergsma

Maintainer: Tim Bergsma <[e-mail]>

References

http://github.com/bergsmat/audited, http://metrumrg.googlecode.com

See Also

as.audited plot.audited as.xlsx.audited metrumrg

Examples

options(audit='Subject')
options(artifact='drop')
rstudiogd()
pc <- data.frame(Theoph)
pc$Subject <- as.numeric(as.character(pc$Subject))
pc <- as.audited(pc,key=c('Subject','Time'))
ex <- pc[pc$Time == 0,,id='ex']
ex$EVID <- 1
pc$EVID <- 0
dat <- alias(id='dat', merge(pc, ex, all=TRUE))
dat <- dat[dat$Subject > 1,]
dat <- dat[dat$Wt > 70,,id='heavier subjects',od='lighter subjects']
audit(dat)
artifact(dat)

# default igraph aesthetics
plot(dat, list())

# unscaled audit aesthetics
plot(dat,scale=FALSE)

# default audit aesthetics
plot(dat)

# adjusting the label presentation
plot(dat, format='%a sub\n%l\n%r row')

# adjusting vertex inflation
plot(dat,inflation=1.2)

# adjusting vertex proportion
plot(dat,proportion=1.2)

# progressively longer audit trails
dat2 <- dat[dat$Subject != 2,,od='subj 2']
dat3 <- dat2[dat2$Subject != 3,,od='subj 3']
dat4 <- dat3[dat3$Subject != 4,,od='subj 4']
dat6 <- dat4[dat4$Subject != 6,,od='subj 6']
dat8 <- dat6[dat6$Subject != 8,,od='subj 8']
dat9 <- dat8[dat8$Subject != 9,,od='subj 9']

# vertex and edge proportions nearly constant across scale
# alternatively, use scale=FALSE and change canvas (image) proportionately
plot(dat)
plot(dat2)
plot(dat3)
plot(dat4)
plot(dat6)
plot(dat8)
plot(dat9)

# a more neutral aesthetic
plot(
  dat3,
  vertex.shape='rectangle',
  vertex.color=NA, # or maybe 'white'
  create.vertex.color=NA,
  drop.vertex.color=NA,
  merge.vertex.color=NA,
  vertex.frame='black',
  edge.color='black',
  vertex.label.color='black',
  merge.edge.color='black',
  drop.edge.color='black'
)

## Not run: 
aud <- audit(dat)
write.audit(aud,'dat.audit')
aud <- read.audit('dat.audit')
as.xlsx(dat,'dat.xlsx')

## End(Not run)

audited documentation built on May 2, 2019, 4:45 p.m.