Should I be afraid of .save and .save1?

[robinsleith]

Was just going to commit the edits to the yaml and R script incorporating truncLen auto etc and encountered the following in the edna-dada2 directory:

rsleith@cfe1 edna-dada2 [master*]$ ls -lthr
total 76K
-rw-rw-r-- 1 rsleith edna 4.8K May  5 16:37 Configurations.md
-rw-rw-r-- 1 rsleith edna 1.7K May  5 16:37 README.md
-rw-rw-r-- 1 rsleith edna 8.7K May  5 16:57 dada2_18S.R
-rw-rw-r-- 1 rsleith edna  429 May  5 16:58 job_submission.sh
-rw-rw-r-- 1 rsleith edna 8.8K May  5 16:59 dada2_pacbio.R
-rw-rw-r-- 1 rsleith edna 1.8K May  5 16:59 dada2_pacbio18S.yaml
-rw-rw-r-- 1 rsleith edna 1.8K May  5 17:00 dada2_pacbio16S.yaml
-rw-rw-r-- 1 rsleith edna 8.7K May  5 17:02 dada2_16S.R
-rw-rw-r-- 1 rsleith edna 1.7K Jun 17 14:47 dada2_18S.yaml
-rw------- 1 rsleith edna    0 Jun 17 19:00 dada2_18S.R.save
-rw-rw-r-- 1 rsleith edna 8.7K Jun 17 19:00 dada2_18S.R.save.1

[btupper]

Hmmm - never seen that before. Let me investigate a bit.

[robinsleith]

Also FWIW dada2_18S.yaml in /mnt/storage/data/edna/packages/edna-dada2 does not match github, meaning changes were made but not committed? Still figuring all this part out.

[btupper]

There is no reference to a ".save" in the dadutils repos.

btupper@cfe1 dadautils [main*]$ git grep -Fn ".save"

It returns nothing, if found it would spit out one line per match. Same is true in edna-dada2

btupper@cfe1 edna-dada2 [master*]$ git grep -Fn ".save"

Which makes me wonder if it is either (a) a user has selected to save session or (b) it is output saved automatically by R when it aborts/crashes. Typically a true session save ends with .Rdata so I'm inclined to think it is (b). But that would mean that the user's working directory was /mnt/storage/data/edna/packages/edna-dada2 which would be weird.

I suggest that I move them to my home directory and then you continue with your work. If they are important we'll still have them. If not we can toss them later. Sound OK?

[robinsleith]

That works!

[btupper]

More...

> load("/mnt/storage/data/edna/packages/edna-dada2/dada2_18S.R.save.1")
Error in load("/mnt/storage/data/edna/packages/edna-dada2/dada2_18S.R.save.1") : 
  bad restore file magic number (file may be corrupted) -- no data loaded
In addition: Warning message:
file ‘dada2_18S.R.save.1’ has magic number 'libra'
  Use of save versions prior to 2 is deprecated 

> z = readRDS("/mnt/storage/data/edna/packages/edna-dada2/dada2_18S.R.save.1")
Error in readRDS("/mnt/storage/data/edna/packages/edna-dada2/dada2_18S.R.save.1") : 
  unknown input format

Ahhhhhaaaaa! It's a text dump of a script. Likely one of our editors did that for some weird reason.

> z = readLines("/mnt/storage/data/edna/packages/edna-dada2/dada2_18S.R.save.1") 
> cat(head(z, n = 10), sep = "\n")
library(dplyr)
library(readr)
library(configr)
library(futile.logger)
library(patchwork)

library(ShortRead)  
library(Biostrings)
library(dada2)

> cat(tail(z, n = 10), sep = "\n")
                               data_path = "data_path",
                               reference_path = "reference_path"),
                             rootname = "global")

if (!interactive()){
  ok <- main(CFG)
  quit(save = "no", status = ok)
}

Anyway - we can delete those.