Hi,
I am asking for your help.
This concerns an error in using Talib in Anaconda Nucleus ( JupyterLab) Notebooks.
When I use LocalHost for my JupyterLab Notebooks, the Talib libraries work just fine. However, I could not import the Talib Library into Anaconda Nucleus ( JupyterLab) Notebooks.
This is part of the Error message I receive:
ValueError: numpy.ndarray size changed, may indicate binary incompatibility.
Expected 96 from C header, got 88 from PyObject
Would love your help and maybe also pointing me in the right direction.
Obviously I am missing something here.
Warmest regards and enjoy your weekends,
Wouter
Hi Wouter,
If the library is working fine on your local system but giving an error in Anaconda Nucleus, it might be due to some discrepancy in the versions. I suggest you try reinstalling the libraries on the Nucleus, this time keeping the library versions the same as on your local system. Please let us know if you still face the issue after doing so.
Hope this helps!
Thanks,
Akshay
HI Akshay,
thank you for wanting to help me.
I realized I needed to do some catching up in terms of Conda knowledge, and so I did first using Anaconda Learning.
Using Local Host, I was able to use the Conda Command Line Manager to create new environments, switch between them, and for now most importantly Fonda install Ta-lib using the instruction line as given at Anaconda.org
To be able to make Ta-lib work I downgraded the local host Python version from 3.11 to 3.10. So now Ta-lib works fine, where I use it in m environment named quanta 
Ok, and now back to Anaconda Cloud/Anaconda Nucleus?( Are both Anaconda Cloud and Anaconda Nucleus basically both used to describe the non local host, thus in the cloud, version of Anaconda?)
So with Anaconda Cloud, I see there is no Conda Command Line manager available? That is correct?
So since there is no Conda Command Line manager available in Anaconda Cloud I tried using the same Conda Command Line commands in the notebook cells instead. ( is that correct way of doing so?)
Using the in my case Jupiter Lab cells to type the same commands as I would do for Conda Command Line Manager, I was getting a long way. However now I got stuck with the following:
When for example trying to conda install --name mnewpackagename, or anything similar, then this is followed by a prompt saying: " Do you wish to proceed? (Y/n)…In local Host you just press "Y" , sill being in the Terminal, and the download begins…
When doing the same in Anaconda Cloud, working with Jupiter Lab, using the cells as Conda Command Line Manager, then thee is no reaction when I would want to confirm the prompt : d you wish to proceed" with Y…
Typing "Y" in a new cell obviously has a complete different meaning and does not give the wanted response, and just typing "Y' on my keyboard in the unknown also does not reach prompt.
Would you be able to help me further with this?
Thanks so much for the help,
Wouter
In JupyterLab, the input cells are primarily designed for executing code, so directly typing "Y" in a new cell won't provide the desired response.
To confirm prompts, you can use the %conda magic command followed by the desired Conda command.
Here's an example:
In a JupyterLab cell, type the following command to execute a Conda command:
%conda install package_name
After executing the cell, if a prompt appears asking, "Do you wish to proceed?", you can confirm it by adding the –yes flag to the command:
%conda install --yes package_name
The –yes flag automatically answers "yes" to any prompts, allowing the installation to proceed without requiring manual confirmation.
By using the %conda magic command with the –yes flag, you should be able to proceed with the desired action without getting stuck on the prompt.
