This will last for a around 15 - 25 seconds or so.
It seems to run whenever i run something that change data. It doesn't matter much if it is a large data structure or just a simple integer. Even deleting a simple integer will trigger this function and will run for 15 seconds. If I run several commands (like several lines) that each change something .rk.get.structure() is run at every command. I don't know about these matters but to me it seems strange that deleting an integer should invlove so much activity.
I have been researching a little about the problem and it seems to get worse the larger the datasets in the workspace. If I delete all data and (as it were) start over with an almost empty workspace it runs fast again. But then as the workspace 'fills up' this function runs for longer and longer.
''
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
I'm experiencing exactly the same problem. For several years I have used Tinn-R under Windows, but now I have decided to turn to Linux, and I've been using Ubuntu for the first time three weeks ago. Rkward seems to me a very good GUI for R, but having to wait 15-25 seconds each time I change data makes it dissapointing. Have your found any solution to this problem since your post?
Thanks
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
> Hello,
> on 2009-01-25 I posted a message in sourceforge, related to
> a problem with .rk.get.structure() in Rkward. I posted that
> message because I saw that you have had exactly the same
> problem that me, i.e., that after reading (or generating)
> some data, the message "R engine busy" appears and lasts for
> 15-25 seconds, while the message .rk.get.structure(....
> appears in the pending jobs window. Could you finally found
> a solution to this problem? I have googled the web and
> nobody else seems to have experienced something similar.
> I would be very grateful if you could help me in this
> question, because I think that R/Rkward is great for
> statistics, but these long waiting times are very disappointing.
>
> Thanks in advance,
> Angelo Santana
Sorry for not seeing your question before.
Basically the problem has largely gone away, I am not sure what exactly caused it though I have some ideas.
It was solved by updating to the newest version which I usually do via svn rather than the package manager. Sometimes, though, I have to first install the official package (to get all the dependencies) then I just install from source on top of the existing installation and things tend to work.
In short: upgrade to newest version. It may seem a bit awkward to have to do it that way but it is my impression that rkward is in a transitional phase right now. Hopefully that will change in the not-too-distant future.
I believe the reason for the lag may be due to some kind of bug in the way rkward adds to R data objects to suits its own structures. This becomes worse when the workspace contains a lot of large objects with many children (like lists and data frames).
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Finally I found the time to investigate this. And ouch, indeed, we have a bug here. The bug is only noticeable, if there are many / complex objects in the user workspace, though. So the short-term "solution" is to clean up the workspace.
This should be fixed in SVN revision 2466 (for the KDE 4 branch, only, not sure, whether I'll backport the changes to the KDE 3 branch). Could you give that a try?
I am seeing this all the time:
My laptop runs at 100% cpu, in the 'pending jobs' window I see this function:
.rk.get.structure(as.environment(".GlobalEnv"),".GlobalEnv",envlevel=-1,namespacename="")
This will last for a around 15 - 25 seconds or so.
It seems to run whenever i run something that change data. It doesn't matter much if it is a large data structure or just a simple integer. Even deleting a simple integer will trigger this function and will run for 15 seconds. If I run several commands (like several lines) that each change something .rk.get.structure() is run at every command. I don't know about these matters but to me it seems strange that deleting an integer should invlove so much activity.
I have been researching a little about the problem and it seems to get worse the larger the datasets in the workspace. If I delete all data and (as it were) start over with an almost empty workspace it runs fast again. But then as the workspace 'fills up' this function runs for longer and longer.
''
Hello,
I'm experiencing exactly the same problem. For several years I have used Tinn-R under Windows, but now I have decided to turn to Linux, and I've been using Ubuntu for the first time three weeks ago. Rkward seems to me a very good GUI for R, but having to wait 15-25 seconds each time I change data makes it dissapointing. Have your found any solution to this problem since your post?
Thanks
> Hello,
> on 2009-01-25 I posted a message in sourceforge, related to
> a problem with .rk.get.structure() in Rkward. I posted that
> message because I saw that you have had exactly the same
> problem that me, i.e., that after reading (or generating)
> some data, the message "R engine busy" appears and lasts for
> 15-25 seconds, while the message .rk.get.structure(....
> appears in the pending jobs window. Could you finally found
> a solution to this problem? I have googled the web and
> nobody else seems to have experienced something similar.
> I would be very grateful if you could help me in this
> question, because I think that R/Rkward is great for
> statistics, but these long waiting times are very disappointing.
>
> Thanks in advance,
> Angelo Santana
Sorry for not seeing your question before.
Basically the problem has largely gone away, I am not sure what exactly caused it though I have some ideas.
It was solved by updating to the newest version which I usually do via svn rather than the package manager. Sometimes, though, I have to first install the official package (to get all the dependencies) then I just install from source on top of the existing installation and things tend to work.
In short: upgrade to newest version. It may seem a bit awkward to have to do it that way but it is my impression that rkward is in a transitional phase right now. Hopefully that will change in the not-too-distant future.
I believe the reason for the lag may be due to some kind of bug in the way rkward adds to R data objects to suits its own structures. This becomes worse when the workspace contains a lot of large objects with many children (like lists and data frames).
Finally I found the time to investigate this. And ouch, indeed, we have a bug here. The bug is only noticeable, if there are many / complex objects in the user workspace, though. So the short-term "solution" is to clean up the workspace.
Bug tracker item:
https://sourceforge.net/forum/message.php?msg_id=6247285
Sorry for the long delay!
Whoops, wrong link. This one leads to the tracker item:
https://sourceforge.net/tracker/?func=detail&aid=2788482&group_id=50231&atid=459007
This should be fixed in SVN revision 2466 (for the KDE 4 branch, only, not sure, whether I'll backport the changes to the KDE 3 branch). Could you give that a try?
Please attach further comments to the bug report:
https://sourceforge.net/tracker/?func=detail&aid=2788482&group_id=50231&atid=459007
Regards
Thomas