Skip to content

Commit

Permalink
minor fixes
Browse files Browse the repository at this point in the history
  • Loading branch information
dipterix committed Jul 8, 2021
1 parent 427e60f commit 60354c8
Show file tree
Hide file tree
Showing 4 changed files with 39 additions and 86 deletions.
18 changes: 18 additions & 0 deletions NEWS.md
Original file line number Diff line number Diff line change
@@ -1,3 +1,21 @@
dipsaus 0.1.6
=======

## Major Changes
* Added `fastcov2` to calculate covariance in parallel
* Added `sumsquared` to calculate sum-squared faster
* Added `sexp_type` to obtain internal data type code

## Bug Fixes
* Robust `get_ram` function. The function will return `NA` on `Solaris`, and actual memories in byte on other platforms.
* Fixed `rs_focus_console` not doing its job
* Fixed `rs_exec` when `wait` is true but not actually waiting

## Enhancements
* Updated `JavaScript` libraries to fix vulnerability issues
* Allow `rs_exec` to return to console after submitting jobs (`RStudio` version 1.4 required)
* Removed unused functions in `README.md`.

dipsaus 0.1.4, 0.1.5
=======

Expand Down
3 changes: 2 additions & 1 deletion R/cpp-fastcov2.R
Original file line number Diff line number Diff line change
Expand Up @@ -84,4 +84,5 @@ fastcov2 <- function(x, y = NULL, col1, col2, df){
df <- nobs - 1
}
fastcov(x, y, nobs, col1, col2, cm1, cm2, df)
}
}

2 changes: 1 addition & 1 deletion R/utils.R
Original file line number Diff line number Diff line change
Expand Up @@ -126,7 +126,7 @@ do_nothing <- function(...){



SEXP_TYPES = list(
SEXP_TYPES <- list(
'0' = c('NILSXP', '0 NILSXP NULL'),
'1' = c('SYMSXP', '1 SYMSXP symbols'),
'2' = c('LISTSXP', '2 LISTSXP pairlists'),
Expand Down
102 changes: 18 additions & 84 deletions cran-comments.md
Original file line number Diff line number Diff line change
@@ -1,95 +1,29 @@
# Current submission (0.1.4.1):
# Current submission (0.1.6):

I apologize for submitting the package so frequently. Thanks Prof. Brian Ripley for pointing out the issues in my code, and I'm trying my best understanding and following the CRAN policy correctly.
1. I changed the my email address as previous one expires soon. As required by `CRAN` policy, I have sent an email from previous address to `CRAN-submissions@r-project.org` explaining the situation. The title is

My previous version showed inconsistency invoking system commands. This was because I was looking for `sysctl` in current environment variable `PATH` instead of `/usr/sbin/` (see the quoted email below).
> Email change for R package `dipsaus` and `threeBrain`
In this update, the function `dipsaus::get_ram` will look for `/usr/sbin/sysctl` when it's not in the path, creating consistent results. If `sysctl` cannot be located, then `NA` will be returned. I also included the command details in the help document.
Please refer to that email.

The function `dipsaus::get_cpu` is defunct, and has been moved to `dipsaus-defunct` list, because this function might produce unpredictable results.
2. I removed `LazyData` specification in `DESCRIPTION` file, as noted on the `CRAN` check website.

3. Self checked locally on `Apple M1 chip`. Online checks include `rhub` (`Linux`, `Solaris`, and `OSX Sierra`) and `Windows` (`old`, `release`, and `devel` provided by `Ligges`). All passed with 1 note as maintainer email has been changed.

```
For dipsaus you did not do as you were asked, and do not look for
sysctl. Without sysctl in the path
> ### Name: get_cpu
> ### Title: Get CPU Chip-set Information
> ### Aliases: get_cpu
>
> ### ** Examples
>
>
> get_cpu()
$vendor_id
[1] NA
$model_name
[1] NA
and with
> example(get_cpu)

get_cp> get_cpu()
$vendor_id
character(0)
attr(,"status")
[1] 1
# Previous submission (0.1.5):

$model_name
[1] "Apple M1"
This and its dependencies have been scheduled for archival on Feb 23.
Do re-read the CRAN policy and do not show contempt for our time.
```

# Previous submission (0.1.4):

This version is to patch an issue caused by improperly invoking system commands without checking their existence.

For your convenience, I paste the original email here:

Version: 0.1.5
Check: LazyData
Result: NOTE
'LazyData' is specified without a 'data' directory
Flavors: r-devel-linux-x86_64-debian-clang, r-devel-linux-x86_64-debian-gcc, r-devel-linux-x86_64-fedora-clang, r-devel-linux-x86_64-fedora-gcc, r-devel-windows-x86_64, r-devel-windows-x86_64-gcc10-UCRT, r-patched-linux-x86_64, r-patched-solaris-x86, r-release-linux-x86_64, r-release-macos-arm64, r-release-macos-x86_64, r-release-windows-ix86+x86_64
Version: 0.1.5
Check: package dependencies
Result: NOTE
Package suggested but not available for checking: ‘RcppRedis’
Flavor: r-patched-solaris-x86
```
On 02/02/2021 08:26, Prof Brian Ripley wrote:
> that is
>
> Rmpi benchmarkme bigparallelr dipsaus disk.frame runjags
>
> On BSD-derived OSes (principally macOS) sysctl is in /usr/sbin or /sbin.
> Those are areas for *s*ystem commands and you cannot assume they are
> on the path (and they probably should not be). So your code needs to
> look for them.
>
> If the runjags message
>
> 'Xgrid is not available on this machine. Xgrid functions are only
> available on machines running Mac OS X 10.5 (Leopard), OS X 10.6 (Snow
> Leopard) or OS X 10.7 (Lion) and with access to an Xgrid controller'
>
> is accurate (and it seems so) this is ancient history and the
> non-functionality should be removed.
>
> Please correct before 2021-02-23 to safely retain your package on CRAN.
>
```

Solution:

* Check if the command `prtconf` (`Solaris`), `sysctl` (`macOS`), `wmic` (`windows`), or `awk` (`linux`) exists on the corresponding operating systems before invoking `system()` calls


Tests:

Tested the updates on `Solaris`, `macOS` (high-sierra with CRAN settings), and `Windows`. The system calls have been properly caught


Additional tests:

* travis (r-release, r-oldrel, and r-devel)
* rhub (windows r-devel, Ubuntu r-release gcc, Fedora r-devel clang)
* win-builder (old, release, devel)
* local macosx (R release)
* local windows (R release, devel)

Check results: 0 error, 0 warning, 0 note

0 comments on commit 60354c8

Please sign in to comment.