Remove an Artifact Given as a md5hash from the Repository

rmFromLocalRepo removes an artifact given as a md5hash from the Repository. To learn more about artifacts visit archivist-package.

rmFromLocalRepo(md5hash, repoDir = aoptions("repoDir"),
  removeData = FALSE, removeMiniature = FALSE, force = FALSE,
  many = FALSE)

rmFromRepo(...)

Arguments

md5hash	A character assigned to the artifact through the use of a cryptographical hash function with MD5 algorithm, or it's abbreviation. This object will be removed. If `many` parameter is set to TRUE then md5hash will be a character vector.
repoDir	A character denoting an existing directory from which an artifact will be removed.
removeData	A logical value denoting whether to remove data along with the `artifact` specified by the `md5hash`. Defualt `FALSE`.
removeMiniature	A logical value denoting whether to remove a miniature along with the `artifact` specified by the `md5hash`. Defualt `FALSE`.
force	A logical value denoting whether to remove data related to more than one artifact. Defualt `FALSE`.
many	A logical value. To accelerate the speed of removing many objects, you can set this parameter to `TRUE` and pass a vector of artifacts' `md5hashes` to a `md5hash` parameter. It is not possible to use a vector of artifacts' `md5hashes` abbreviations - see `Note`. By default, set to `FALSE`.
...	All arguments are being passed to `rmFromLocalRepo`.

Details

rmFromLocalRepo removes an artifact given as a md5hash from the Repository. To be more precise, an artifact is removed both from backpack.db file(the SQLite database)and gallery subdirectory, where the artifacts are stored as md5hash.rda files.

Important: instead of giving the whole md5hash character, a user can simply give its first few characters. For example, "a09dd" instead of "a09ddjdkf9kj33dcjdnfjgos9jd9jkcv". All artifacts with the same md5hash abbreviation will be removed from the Repository.

rmFromLocalRepo provides functionality that enables us to delete miniatures of the artifacts (.txt or .png files) while removing .rda files. To delete miniature of the artifact use removeMiniature = TRUE argument. Moreover, if the data from the artifact is archived then there is a possibility to delete this data while removing the artifact. Simply use removeData = TRUE argument.

If one wants to remove all artifacts created between two dates, it is suggested to perform:

obj2rm <- searchInLocalRepo( tag = list(dateFrom, dateTo), repoDir = )
sapply(obj2rm, rmFromLocalRepo, repoDir = )

Note

md5hash can be a result of the searchInLocalRepo function called by tag = NAME argument, where NAME is a Tag that describes the property of the artifacts to be deleted.

It is not possible to use a vector of artifacts' md5hashes abbreviations while using many = TRUE argument. This assumption was made to protect a user from removing, by accident, too many artifacts from the Repository.

For more information about Tags check Tags.

Contact

Bug reports and feature requests can be sent to https://github.com/pbiecek/archivist/issues

References

Biecek P and Kosinski M (2017). "archivist: An R Package for Managing, Recording and Restoring Data Analysis Results." _Journal of Statistical Software_, *82*(11), pp. 1-28. doi: 10.18637/jss.v082.i11 (URL: http://doi.org/10.18637/jss.v082.i11). URL https://github.com/pbiecek/archivist

Examples

# NOT RUN {
# objects preparation
data.frame object
data(iris)

# ggplot/gg object
library(ggplot2)
df <- data.frame(gp = factor(rep(letters[1:3], each = 10)),y = rnorm(30))
library(plyr)
ds <- ddply(df, .(gp), summarise, mean = mean(y), sd = sd(y))
myplot123 <- ggplot(df, aes(x = gp, y = y)) +
  geom_point() +  geom_point(data = ds, aes(y = mean),
                             colour = 'red', size = 3)

# lm object
model <- lm(Sepal.Length~ Sepal.Width + Petal.Length + Petal.Width, data= iris)
model2 <- lm(Sepal.Length~ Sepal.Width + Petal.Width, data= iris)
model3 <- lm(Sepal.Length~ Sepal.Width, data= iris)

# agnes (twins) object
library(cluster)
data(votes.repub)
agn1 <- agnes(votes.repub, metric = "manhattan", stand = TRUE)

# fanny (partition) object
x <- rbind(cbind(rnorm(10, 0, 0.5), rnorm(10, 0, 0.5)),
           cbind(rnorm(15, 5, 0.5), rnorm(15, 5, 0.5)),
           cbind(rnorm( 3,3.2,0.5), rnorm( 3,3.2,0.5)))
fannyx <- fanny(x, 2)

# creating example Repository - on which examples will work

exampleRepoDir <- tempfile()
createLocalRepo(repoDir = exampleRepoDir)
myplot123Md5hash <- saveToLocalRepo(myplot123, repoDir=exampleRepoDir)
irisMd5hash <- saveToLocalRepo(iris, repoDir=exampleRepoDir)
modelMd5hash  <- saveToLocalRepo(model, repoDir=exampleRepoDir)
agn1Md5hash <- saveToLocalRepo(agn1, repoDir=exampleRepoDir)
fannyxMd5hash <- saveToLocalRepo(fannyx, repoDir=exampleRepoDir)

# let's see how the Repository looks like: show
showLocalRepo(method = "md5hashes", repoDir = exampleRepoDir)
showLocalRepo(method = "tags", repoDir = exampleRepoDir)

# let's see how the Repository looks like: summary
summaryLocalRepo( exampleRepoDir )

# remove examples

rmFromLocalRepo(fannyxMd5hash, repoDir = exampleRepoDir)
# removeData = FALSE default argument provides from removing archived
# fannyxMd5hash object's data from the Repository and the gallery
rmFromLocalRepo(irisMd5hash, repoDir = exampleRepoDir)
 
# let's see how the Repository looks like: show
showLocalRepo(method = "md5hashes", repoDir = exampleRepoDir)
showLocalRepo(method = "tags", repoDir = exampleRepoDir)

# let's see how the Repository looks like: summary
summaryLocalRepo( exampleRepoDir )


# one can have the same object archived three different times,
# there will appear a warning message
agn1Md5hash2 <- saveToLocalRepo(agn1, repoDir=exampleRepoDir)
agn1Md5hash3 <- saveToLocalRepo(agn1, repoDir=exampleRepoDir)

# md5hashes are the same for the same object (agn1)
agn1Md5hash == agn1Md5hash2
agn1Md5hash2 == agn1Md5hash3

# but in the Repository database (backpack.db)
# there are three identical rows describing the object
# as well as three identical rows describing object's data.

# let's see how the Repository looks like: show
showLocalRepo(method = "md5hashes", repoDir = exampleRepoDir)
showLocalRepo(method = "tags", repoDir = exampleRepoDir)

# let's see how the Repository looks like: summary
summaryLocalRepo( exampleRepoDir )
# in spite of multiplying object's appearance in database it is

# one easy call removes them all but this call will result in error
rmFromLocalRepo(agn1Md5hash, repoDir = exampleRepoDir, removeData = TRUE, 
            removeMiniature = TRUE)

# soultion to that is
rmFromLocalRepo(agn1Md5hash, repoDir = exampleRepoDir, removeData = TRUE, 
            removeMiniature = TRUE, force = TRUE)
# removeMiniature = TRUE removes miniatures from the gallery folder

# rest of the artifacts can be removed for example by
# looking for dates of creation and then removing all objects
# created in a specific period of time

obj2rm <- searchInLocalRepo( pattern = list(dateFrom = Sys.Date(), dateTo = Sys.Date()),
                             repoDir = exampleRepoDir )
sapply(obj2rm, rmFromLocalRepo, repoDir = exampleRepoDir)
# above function call removed all objects which were created in these examples.
# Note that in the gallery folder there may be still some miniatures as
# removeMiniature parameter is set to FALSE

# let's see how the Repository looks like: show
showLocalRepo(method = "md5hashes", repoDir = exampleRepoDir)
showLocalRepo(method = "tags", repoDir = exampleRepoDir)

# one can also delete objects of a specific class
modelMd5hash  <- saveToLocalRepo(model, repoDir=exampleRepoDir)
model2Md5hash  <- saveToLocalRepo(model2, repoDir=exampleRepoDir)
model3Md5hash  <- saveToLocalRepo(model3, repoDir=exampleRepoDir)
showLocalRepo(method = "md5hashes", repoDir = exampleRepoDir)

objMd5hash <- searchInLocalRepo("class:lm", repoDir = exampleRepoDir)
sapply(objMd5hash, rmFromLocalRepo, repoDir = exampleRepoDir, removeData = TRUE, force = TRUE)
showLocalRepo(method = "md5hashes", repoDir = exampleRepoDir)
summaryLocalRepo( exampleRepoDir )


# one can remove object specifying only its md5hash abbreviation
(myplo123Md5hash <- saveToLocalRepo(myplot123, repoDir=exampleRepoDir))
showLocalRepo(method = "md5hashes", repoDir = exampleRepoDir)

# If md5hash is "db50a4e667581f8c531acd78ad24bfee" then
# model abbreviation might be : "db50a"
# Note that with each evaluation of createEmptyRepo function new md5hashes
# are created. This is why, in your evaluation of the code, artifact 
# myplo123Md5hash will have a different md5hash and the following
# instruction will result in an error.
rmFromLocalRepo("db40a", repoDir = exampleRepoDir, removeData = TRUE)
summaryLocalRepo( repoDir = exampleRepoDir )


# removing an example Repository

deleteLocalRepo( exampleRepoDir, TRUE)


######
######
REMOVING MANY ARTIFACTS
######
######

data(iris)

# lm object
model <- lm(Sepal.Length~ Sepal.Width + Petal.Length + Petal.Width, data= iris)

# agnes (twins) object
library(cluster)
data(votes.repub)
agn1 <- agnes(votes.repub, metric = "manhattan", stand = TRUE)

# fanny (partition) object
x <- rbind(cbind(rnorm(10, 0, 0.5), rnorm(10, 0, 0.5)),
           cbind(rnorm(15, 5, 0.5), rnorm(15, 5, 0.5)),
           cbind(rnorm( 3,3.2,0.5), rnorm( 3,3.2,0.5)))
fannyx <- fanny(x, 2)

# lda object
library(MASS)

Iris <- data.frame(rbind(iris3[,,1], iris3[,,2], iris3[,,3]),
                  Sp = rep(c("s","c","v"), rep(50,3)))
train <- c(8,83,115,118,146,82,76,9,70,139,85,59,78,143,68,
           134,148,12,141,101,144,114,41,95,61,128,2,42,37,
           29,77,20,44,98,74,32,27,11,49,52,111,55,48,33,38,
           113,126,24,104,3,66,81,31,39,26,123,18,108,73,50,
           56,54,65,135,84,112,131,60,102,14,120,117,53,138,5)
lda1 <- lda(Sp ~ ., Iris, prior = c(1,1,1)/3, subset = train)

# qda object
tr <- c(7,38,47,43,20,37,44,22,46,49,50,19,4,32,12,29,27,34,2,1,17,13,3,35,36)
train <- rbind(iris3[tr,,1], iris3[tr,,2], iris3[tr,,3])
cl <- factor(c(rep("s",25), rep("c",25), rep("v",25)))
qda1 <- qda(train, cl)

# glmnet object
library( glmnet )

zk=matrix(rnorm(100*20),100,20)
bk=rnorm(100)
glmnet1=glmnet(zk,bk)
 
# Creating example Repository so that we may see it on our computer

exampleRepoDir <- tempfile()
createLocalRepo( repoDir = exampleRepoDir, force = TRUE)
saveToLocalRepo( iris, repoDir=exampleRepoDir)
saveToLocalRepo( model, repoDir=exampleRepoDir )
saveToLocalRepo( agn1, repoDir=exampleRepoDir )
saveToLocalRepo( fannyx, repoDir=exampleRepoDir )
saveToLocalRepo( lda1, repoDir=exampleRepoDir )
saveToLocalRepo( glmnet1, repoDir=exampleRepoDir )

ArtifactsAndData <- unique(showLocalRepo(repoDir = exampleRepoDir)[,1])
ArtifactsData <- unique(searchInLocalRepo(pattern = "relationWith", fixed = FALSE,
                                   repoDir = exampleRepoDir))
Artifacts <- setdiff(ArtifactsAndData, ArtifactsData)


# Removing many artifacts with many = TRUE argument
rmFromLocalRepo(Artifacts, repoDir = exampleRepoDir, many = TRUE)

# We may notice, in two ways, that artifacts' data is still in "exampleRepoDir".
# Either we may look into gallery folder of "exampleRepoDir" 
list.files(file.path(exampleRepoDir, "gallery"))
# or show how database.db file looks like.
showLocalRepo(repoDir = exampleRepoDir) # artifacts' data is there indeed!

# If we want to remove artifact's data now we simply call rmFromLocalRepo function
# with removeData = TRUE additional argument.
rmFromLocalRepo(Artifacts, repoDir = exampleRepoDir, removeData = TRUE,  many = TRUE)

# We receive a warning as Artifacts are no longer in the repository. 
# However, let's check what happened with Artifact's data.
showLocalRepo(repoDir = exampleRepoDir) # They were removed.
# Perhaps you may think that "exampleRepoDir" is empty as database indicates. However,
# if you look into gallery folder there will be some ".txt" or ".png" files. 
list.files(file.path(exampleRepoDir, "gallery"))

# Those are probably, the so called, Miniatures. Let's try to remove them.
# In order to do it we call rmFromLocalRepo function with removeMiniature = TRUE argument.
rmFromLocalRepo(Artifacts, many = TRUE, repoDir = exampleRepoDir, removeMiniature = TRUE)

# Again we receive a warning as Artifacts are no longer in the repository but ...
list.files(file.path(exampleRepoDir, "gallery")) 
# gallery folder is empty now! Artifact's miniature's were removed.


# Of course we may have done all these instructions by one simple function call.
# rmFromLocalRepo(Artifacts, many = TRUE, repoDir = exampleRepoDir,
#            removeData = TRUE, removeMiniature = TRUE)
# Nevertheless, it may be instructive to see how it is done step by step.

# removing an example Repository
deleteLocalRepo(repoDir = exampleRepoDir, deleteRoot = TRUE)
   
rm( exampleRepoDir )
# }