These are chat archives for spyder-ide/public

25th
Oct 2018
aperchak
@aperchak
Oct 25 2018 00:29
Thanks for the suggestions on github. Just wanted to say that it is evening here and I will be continuing tomorrow, but I do in fact have the program "Amplitube" installed as I am an audio engineer in my profession. It could indeed be the culprit, I will report tomorrow.
CAM Gerlach
@CAM-Gerlach
Oct 25 2018 00:32
Okay, thanks—I should have checked earlier. Its virtually certain to be the culprit, based on other reports, due to its bad practices of installing dlls directly into the System32 directory, which can mess with all other applications. Before you uninstall it, please do post your full screenshot and the output of the two where commands in order to help our troubleshooting for others if you could. Thanks!
aperchak
@aperchak
Oct 25 2018 00:35
No problem, will do.
CAM Gerlach
@CAM-Gerlach
Oct 25 2018 00:48
Thanks!
Md. Ariful Islam
@Arif-IITKGP
Oct 25 2018 05:09
Whenever I work in spyder, I face a problem of memory out.
Please help me.
It becomes very slow and not able to do anything in the system.
CAM Gerlach
@CAM-Gerlach
Oct 25 2018 05:11
For us to be able to do that other than give generic troubleshooting advice, you're going to need to give us specific details of the behavior you are experiencing and the exact text of any error messages.
Several significant memory leaks have been fixed and performance optimizations implemented to greatly reduce memory consumption of Spyder's variable explorer in recent versions. If you don't have Spyder 3.3.1, you should upgrade (e.g. with conda update spyder) and see if the problem reoccurs. You can also try refreshing your Spyder configuration (with spyder --reset) in case any of the logs or caches, etc. is taking up too much memory.
Md. Ariful Islam
@Arif-IITKGP
Oct 25 2018 05:18

Traceback (most recent call last):

File "<ipython-input-5-9fb7cc023da4>", line 1, in <module>
get_ipython().kernel.get_value('Data_All')

File "C:\Users\user\Anaconda2\lib\site-packages\spyder\utils\ipython\spyder_kernel.py", line 140, in get_value
publish_data({'spy_data': value})

File "C:\Users\user\Anaconda2\lib\site-packages\ipykernel\datapub.py", line 62, in publish_data
ZMQInteractiveShell.instance().data_pub.publish_data(data)

File "C:\Users\user\Anaconda2\lib\site-packages\ipykernel\datapub.py", line 40, in publish_data
item_threshold=session.item_threshold,

File "C:\Users\user\Anaconda2\lib\site-packages\ipykernel\serialize.py", line 93, in serialize_object
buffers.insert(0, pickle.dumps(cobj, PICKLE_PROTOCOL))

MemoryError: out of memory

CAM Gerlach
@CAM-Gerlach
Oct 25 2018 05:23
As I mentioned previoulsy, what version of Spyder are you using? On what operating system version? When exactly does the error occur (what specific operation were you attempting)? How much RAM does your machine have, and what sizes of datasets are you working with? What version of IPython and ipykernel are you using? Make sure to update spyder-kernels as well when updating Spyder.
Md. Ariful Islam
@Arif-IITKGP
Oct 25 2018 05:30
Spyder 3.1.4
CAM Gerlach
@CAM-Gerlach
Oct 25 2018 05:30
!!!
Md. Ariful Islam
@Arif-IITKGP
Oct 25 2018 05:30
Windows 7
RAM: 8GB
Data Size: 1.8 GB
CAM Gerlach
@CAM-Gerlach
Oct 25 2018 05:31
That's over 1.5 years old. The current version is Spyder 3.3.1; upgrade with conda update spyder.
Md. Ariful Islam
@Arif-IITKGP
Oct 25 2018 05:32
Operation causing error: Just double click on the data array in variable explorer
It is being updated
CAM Gerlach
@CAM-Gerlach
Oct 25 2018 05:32
It is being updated
Do you mean you are updating it right now?
Md. Ariful Islam
@Arif-IITKGP
Oct 25 2018 05:32
yes sir
Thank you very much sir
CAM Gerlach
@CAM-Gerlach
Oct 25 2018 05:35
Excellent. Very likely it is due to the memory leaks and un-optimized performance which was fixed a few versions ago. Although, if it is a very large array (~100s of MB-1GB +), you might just need more RAM or close anything else using up memory, since 8 GB is on the low side for a modern data science workstation.
Md. Ariful Islam
@Arif-IITKGP
Oct 25 2018 05:37
Okay, I got it.
Thanks for your help.
I will revert to you if the problem persists.
CAM Gerlach
@CAM-Gerlach
Oct 25 2018 05:58
Best of luck!
raphaelquast
@raphaelquast
Oct 25 2018 07:41
@CAM-Gerlach thanks for the great suggestion! I wasn't aware that 2 separate spyder versions are supported via virtual conda environments. I'll definitively give it a try since i've been using spyder now since quite some time and there are some features coming up that i really appreciate...
CAM Gerlach
@CAM-Gerlach
Oct 25 2018 07:43

I wasn't aware that 2 separate spyder versions are supported via virtual conda environments

In fact, you can create any number of conda environments (not to be confused with virtualenvs, which have generally equivalent functionality but are a distinct system) to install any equivalent number of different versions of packages; that, in fact, is the whole point of such environments in the first place.

@raphaelquast Thanks for trying out the development version, and let us know if you find any bugs or issues (if they aren't already fixed in the upcoming Beta 2). We hope you enjoy all the new enhancements!
raphaelquast
@raphaelquast
Oct 25 2018 11:39
@CAM-Gerlach yep, i know that, but i was not sure if spyder is in fact contained within each environment or just as a global editor whose console is connected to the chosen environment (i.e. if messing around with my sypder installation in one environment might affect others) good to know that this is not the case :-)
CAM Gerlach
@CAM-Gerlach
Oct 25 2018 17:57

in fact contained within each environment or just as a global editor whose console is connected to the chosen environment

Spyder is a normal Python package you can install in any Python environment, but with the changes in 3.3.0/3.3.1, you can install Spyder in one environment and then connect its console to any conda env, virtualenv, other Python install on your computer or even a remote machine, so you only need to install it once to work anywhere.

aperchak
@aperchak
Oct 25 2018 23:26
Hi @CAM-Gerlach, pardon my novice question - you asked for the output of the where mkl_core.dll and where mkl_intel_thread.dll command in the Anaconda prompt, correct?
CAM Gerlach
@CAM-Gerlach
Oct 25 2018 23:47
Yes, exactly right @aperchak , ideally done both before and after you remove amplitute. If you could post that in the thread along with a screenshot of the actual error message, that would be great. Thanks!