R check complaints - Githubissues

mroughan / LinePicking

Numerical code for geometric probability problems, in particular density functions for the "line picking" problem.

0 stars 1 forks source link

R check complaints #17

Closed jonotuke closed 12 years ago

jonotuke commented 12 years ago

The log for R CMD check gives this

checking compiled code ... NOTE File '/private/var/folders/rn/7qtz5j6548j7yx1_sh7lqknr0000gr/T/RtmpxPVsKN/LinePicking.Rcheck/LinePicking/libs/x86_64/LinePicking.so': Found '___stderrp', possibly from 'stderr' (C) Object: 'beta.o' Found '_exit', possibly from 'exit' (C) Object: 'beta.o'

Compiled code should not call functions which might terminate R nor write to stdout/stderr instead of to the console.

Is it possible to change these?

lamestllama commented 12 years ago

Jono I will investigate now

mroughan commented 12 years ago

fprintf's should be replaced by the macro, or a new one for writing to stderr

On 29/08/12 13:40, Eric Parsonage wrote:

Jono I will investigate now

— Reply to this email directly or view it on GitHub https://github.com/mroughan/LinePicking/issues/17#issuecomment-8114774.

lamestllama commented 12 years ago

I am thinking we also need to return a status so that we know the function has failed

On 29/08/2012, at 1:52 PM, Matthew Roughan wrote:

fprintf's should be replaced by the macro, or a new one for writing to stderr

On 29/08/12 13:40, Eric Parsonage wrote:

Jono I will investigate now

— Reply to this email directly or view it on GitHub https://github.com/mroughan/LinePicking/issues/17#issuecomment-8114774.

— Reply to this email directly or view it on GitHub.

mroughan commented 12 years ago

R doesn't like functions that return values.

On 29/08/12 13:55, Eric Parsonage wrote:

I am thinking we also need to return a status so that we know the function has failed

On 29/08/2012, at 1:52 PM, Matthew Roughan wrote:

fprintf's should be replaced by the macro, or a new one for writing to stderr

On 29/08/12 13:40, Eric Parsonage wrote:

Jono I will investigate now

— Reply to this email directly or view it on GitHub

https://github.com/mroughan/LinePicking/issues/17#issuecomment-8114774.

— Reply to this email directly or view it on GitHub.

— Reply to this email directly or view it on GitHub https://github.com/mroughan/LinePicking/issues/17#issuecomment-8114963.

lamestllama commented 12 years ago

I return it by reference

On 29/08/2012, at 2:00 PM, Matthew Roughan wrote:

R doesn't like functions that return values.

On 29/08/12 13:55, Eric Parsonage wrote:

I am thinking we also need to return a status so that we know the function has failed

On 29/08/2012, at 1:52 PM, Matthew Roughan wrote:

fprintf's should be replaced by the macro, or a new one for writing to stderr

On 29/08/12 13:40, Eric Parsonage wrote:

Jono I will investigate now

— Reply to this email directly or view it on GitHub

https://github.com/mroughan/LinePicking/issues/17#issuecomment-8114774.

— Reply to this email directly or view it on GitHub.

— Reply to this email directly or view it on GitHub https://github.com/mroughan/LinePicking/issues/17#issuecomment-8114963.

— Reply to this email directly or view it on GitHub.

lamestllama commented 12 years ago

Jono

I have added an extra parameter thus

/* regularized incomplete beta function / double beta_inc(double, double, double, int result);

pass it an integer by reference if after the call it is non zero then the function failed because one of the three inputs where out of the alowed range

if you get 1 it was the first input if you get 2 it was the second and if you get 3 well you get the idea.

This should fix the problem

Eric

On 29/08/2012, at 1:35 PM, jonotuke wrote:

The log for R CMD check gives this