tazovsky / cachemer Goto Github PK

Add summary/print method to print cached data. Use data.table or maybe tibble.

Fix error:

Błąd w poleceniu 'names(qte.list) <- names(res)':
atrybut 'names' [3] musi mieć tę samą długość co wektor [2]

Code to reproduce error:

dir.create(tmp.dir <- tempfile())
config.file <- file.path(tmp.dir, "config.yaml")
cache <- cachemer$new(path = config.file)

testFun <- function(a, b) {
  (a+b) ^ (a*b)
}

cache$setLogger(TRUE)

result1 %c-% testFun(2, 3)

Doesn't work:

vec.arg <- 1:100
list.arg <- list(d = 2, e = 3)
res %c-% testFun(a = vec.arg, 1, list.arg)

as.character(target)[1] is not good for parsing variable name to string because it causes assignement to $:

Browse[3]> target
state$active_covariates
Browse[3]> as.character(target)[1]
[1] "$"

Create pipe to cache LHS:

res <- a %>% sin %>% sum %c% . - in this case LHS is a %>% sin %>% sum

Add readme containing:

• Package description
• Installation tutorial
• How to use
• Some use cases
• Microbenchmark

• Add method to save current cache to file
• Add method to restore cache from file
• Investigate what would be the best output format for saving cache
• Use future to save cache in background and do not block code

Create method to get cached object by:

• hash
• function name
• output value (conditions: %in%, <=, >=?)

HINT: Create data.table object from cached objects to filter fast and easy.

fst R package seems to be faster than data.table or feather: http://www.fstpackage.org/ - maybe it's worth considering to use it to save cache?

Collect statistics about the use of functions, arguments, parent functions, etc.

• Add CI to project (Travis?)
• Test coverage badge
• Build badge
• Download per month badge

Branch: https://github.com/Tazovsky/cachemeR/tree/feature/add-reset-cache