I ran into this behaviour when accidentally running a line of code that I shouldn't have. When saving over an rds with an object that's not found, I would have expected saveRDS to not touch the file. saveRDS(iris, "test.rds") file.size("test.rds") #> [1] 1080 saveRDS(no_object_here, "test.rds") #> Error in saveRDS(no_object_here, "test.rds"): object 'no_object_here' not found file.size("test.rds") #> [1] 20 file.remove("test.rds") [[alternative HTML version deleted]]
Tomas Kalibera
2018-Feb-08 14:18 UTC
[Rd] saveRDS() overwrites file when object is not found
Thanks, this has been already reported as bug 17358. Addressed in R-devel 74238. R may still create a corrupt file, though, in other circumstances (e.g. if it runs out of memory or is interrupted during serialization, etc). Tomas On 02/07/2018 04:14 PM, Kenny Bell wrote:> I ran into this behaviour when accidentally running a line of code that I > shouldn't have. > > When saving over an rds with an object that's not found, I would have > expected saveRDS to not touch the file. > > saveRDS(iris, "test.rds") > file.size("test.rds") > #> [1] 1080 > saveRDS(no_object_here, "test.rds") > #> Error in saveRDS(no_object_here, "test.rds"): object 'no_object_here' > not found > file.size("test.rds") > #> [1] 20 > file.remove("test.rds") > > [[alternative HTML version deleted]] > > ______________________________________________ > R-devel at r-project.org mailing list > https://stat.ethz.ch/mailman/listinfo/r-devel
Seemingly Similar Threads
- [FORGED] recordPlot/replayPlot not working with saveRDS/readRDS
- recordPlot/replayPlot not working with saveRDS/readRDS
- Risk of readRDS() not detecting race conditions with parallel saveRDS()?
- readRDS and saveRDS
- Model object, when generated in a function, saves entire environment when saved