_From [email protected] on 2013-12-10T22:21:42Z_
Is it possible to list only user-created variables in the explorer? If the "Exclude unsupported data types" tick is checked, I don't see objects like, for example, an xaml.objectify output, while if this is tick unchecked, there are tons of other symbols making the explorer really hard to use. I'm a newbie to Python so I don't know how hard it is to implement the filter I want but at least I think an acceptable and implementable solution would be to keep a list of symbols that were created on startup or by subsequent import calls.
Spyder 2.2.4, Python 2.7.5, Windows 7 x64
_Original issue: http://code.google.com/p/spyderlib/issues/detail?id=1676_
_From ccordoba12 on 2013-12-11T17:20:05Z_
It won't be so easy but you are making a good point here. We'll see what we can do about it.
Labels: Cat-VariableExplorer MS-v2.4
Hi !
I know this is a bit of an old request, but I do agree with Peter. A filter would be really a plus, as long as user's classes cannot be accepted as 'supported'. I intensively use the variable explorer and create my own classes, and have the 'Exclude unsupported data types' un-ticked literally overcharge the explorer.
We're working to improve this for Spyder 4.0 (which we hope to release in June/July 2017)
@ccordoba12 Ever since I started with Spyder I've been planning to implement a "Hide functions and modules" filter on 3.x, particularly on that branch now with the changes to the cloudpickle and the object explorer allowing more objects, e.g. modules, to be supported and viewable, etc. It shouldn't be too terribly tough to do from what I can tell, just requires adding the fairly simple logic, then changing a few different things a few different places for the frontend part. Has any substantial work already been done on this, as I don't want to duplicate effort? I don't see it at least on the GUI off a master checkout... let me know how'd you like me to proceed.
Of course, I don't get to it right away since I have the FAQ to polish, several bugs to fix and issues to report, other smaller things to do, etc, but its certainly in the queue.
Has any substantial work already been done on this, as I don't want to duplicate effort?
Nope, nothing has been done about this.
Of course, I don't get to it right away since I have the FAQ to polish, several bugs to fix and issues to report, other smaller things to do, etc, but its certainly in the queue.
Don't worry about it, but I don't want to see more new features in 3.x, because it means we have to keep maintaining them in that branch. My hope is to release three, maybe four more 3.2.x versions, then focus entirely on Spyder 4.
The cloudpickle thing was necessary because a future, backwards incompatible release of ipykernel would have broken our Variable Explorer.
The problem was that we were using a deprecated API, but it was deprecated after I released Spyder 3.0 and I was unsure how to fix that.
Don't worry about it, but I don't want to see more new features in 3.x, because it means we have to keep maintaining them in that branch. My hope is to release three, maybe four more 3.2.x versions, then focus entirely on Spyder 4.
Okay, cool. I'll base against master then when I get to it.
Any news on this? I'm using spyder for the first time and there is a ton of numpy classes and objects flooding the explorer mixing with my stuff. I'd really love a way to hide them...
Thanks!
@danizil "Exclude callables and modules" (essentially what I proposed above) is now an option in Spyder 4's Variable Explorer, so does that solve your issue? That should exclude non-user-created variables (as well as user-created functions, but those rarely ever need to be introspected directly), at least as well as we programmatically can detect. You can also try "Exclude unsupported data types" if you are only interested in things you can easily inspect with the Variable Explorer. Does that resolve this? If so this issue can probably be closed @ccordoba12
Agreed. It's really not possible to differentiate between your own data and data structures, and those you imported from other libraries, sorry.
Hi, It will be really great if I can only see the the variables that I generated from my script on the explorer. I tried all the excluding filters but none worked. Is there any solution currently on that? It is really difficult to work with all these extra variables that I do not need for my analysis.
That's not technically possible, sorry.
It is really difficult to work with all these extra variables that I do not need for my analysis.
Then don't import them in the global namespace.
It is really difficult to work with all these extra variables that I do not need for my analysis.
Then don't import them in the global namespace.
great, sorted!