bugGNU Octave - Bugs: bug #47063, computation between single and...

 
 

bug #47063: computation between single and double performed with single precision

Submitter:  CarnĂ« Draug <carandraug>
Submitted:  Fri 05 Feb 2016 02:45:01 AM UTC
   
 
Category:  Interpreter Severity:  3 - Normal
Priority:  3 - Low Item Group:  WTF, Matlab?!?
Status:  Confirmed Assigned to:  None
Originator Name:  Open/Closed:  * Open
Release:  * dev Operating System:  * Any
Fixed Release:  None Planned Release:  None
* Mandatory Fields

Add a New Comment Rich Markup
   

Fri 18 Mar 2016 02:43:58 PM UTC, comment #4: 

Since you asked on IRC...

It's better to write something like "perform mixed single-double ops in double precision (bug #...)" as the summary instead of "fix the bug...".  Many patches are bug fixes and you're already tagging the bug report number so it's redundant to say "fix the bug".

Similarly, just write

  * foo.h, foo.cc (function_name): New function.

instead of "add function ...".  I think there are some tips about this in the Octave style guidelines in the Octave manual and also in the GNU coding standards document that it references.

John W. Eaton <jwe>
Group administrator
Fri 18 Mar 2016 02:25:26 PM UTC, comment #3: 

You are on the right track, but "demotion" means to reduce the rank of something, so that seems like the wrong word for a function that promotes float to double.  It also makes no sense to have a demotion or promotion function that just makes a new value of the same type.  Instead, if we are going to promote singles to doubles, we should remove the function that used to demote doubles to singles and add a promotion function for singles to doubles.  And rename all of them, if we only need to do promotion.  And you need to do it for matrix types as well, not just scalars.

But I'm still not sure this is a good idea.  I'm not sure we should follow Matlab in this case.

John W. Eaton <jwe>
Group administrator
Fri 18 Mar 2016 02:15:22 PM UTC, comment #2: 

I have attached a file(patch) that solves this problem for single-double operations.
The idea is preventing double variables from being casted to single and on the other hand casting singles to doubles in order to perform the operation smoothly.

Amr Mohamed <amr_keleg>
Thu 10 Mar 2016 10:47:11 AM UTC, comment #1: 

I would like towork on this bug can anyone guide me how interpreter works in octave.

SatyamPandey <spandey>
Fri 05 Feb 2016 02:45:01 AM UTC, original submission:  

An operation between a double and a single returns a single for Matlab compatibility. However, it seems that Matlab actually performs the computations with doubles. See https://stackoverflow.com/questions/35213845/subtracting-double-precision-from-single-precision-gives-me-0-not-what-i-want


## in octave
octave> single (sin (10)) - sin (10)
ans = 0
octave> single (double (single (sin (10))) - sin (10))
ans =   -1.8719e-08

----------

## in matlab
>> single (sin (10)) - sin (10)

ans =

 -1.8719e-08


Carnë Draug <carandraug>
Group Member

 

(Note: upload size limit is set to 16384 kB, after insertion of the required escape characters.)

Attach Files:
   
   
Comment:
   

Attached Files
file #36686:  ScalarFloatBugFix.diff added by amr_keleg (2KiB - text/x-patch - Kindly Review the attached patch)

 

Depends on the following items: None found

Items that depend on this one: None found

 

Carbon-Copy List
  • -email is unavailable- added by rik5 (Updated the item)
  • -email is unavailable- added by amr_keleg (Updated the item)
  • -email is unavailable- added by spandey (Posted a comment)
  • -email is unavailable- added by jwe (Updated the item)
  • -email is unavailable- added by carandraug (Submitted the item)
  •  

    There are 0 votes so far. Votes easily highlight which items people would like to see resolved in priority, independently of the priority of the item set by tracker managers.

    Only group members can vote.

     

    Follow 7 latest changes.

    Date Changed by Updated Field Previous Value => Replaced by
    2019-06-08 mtmiller Carbon-CopyRemoved 80942 -
    2019-06-08 rik5 Priority5 - Normal 3 - Low
        StatusNone Confirmed
    2016-03-18 amr_keleg Attached File- Added ScalarFloatBugFix.diff, #36686
    2016-02-21 mtmiller CategoryNone Interpreter
        Operating SystemGNU/Linux Any
    2016-02-05 jwe Item GroupInaccurate Result WTF, Matlab?!?

    Back to the top

    Powered by Savane 3.13-4448.
    Corresponding source code