-
Error with setting up libraries
I am trying to set up the machine learning package and am facing an error. I will summarise my steps below and then share the error message.
1. Installed PYKX (using pip) and license is confirmed.
2. Followed the steps for embedPy and installed by downloading a release from the archives and moving p.q, p.k, and p.dll files to my QHOME directory (as instructed).
3. Cloned the ml github repository into my local repository.
4. Started a q process that loads the ml.q file. (ml/ml/ml.q)
I get the following error suggesting that I failed to load the module:
“””
(.venv) PS C:\Users\ali\OneDrive – Onyx Commodities\Documents\Projects\kdbq environment> q main.q
KDB+ 4.1 2024.10.16 Copyright (C) 1993-2024 Kx Systems
w64/ 16(24)core 65219MB ali onyx-uk-00071 10.9.30.57 EXPIRE 2026.01.08 a.li2001@hotmail.com KDB INSIGHTS #5023662
‘Failed to load PyKX or embedPy with error: p.q. OS reports: The system cannot find the file specified.
[7] C:\q\ml\ml.q:12: {‘”Failed to load PyKX or embedPy with error: “,x}
^
[6] C:\q\ml\ml.q:11: {@[{system”l “,x;.pykx.loaded:0b};”p.q”;
^
{‘”Failed to load PyKX or embedPy with error: “,x}]}
q.ml))’2025.01.09T03:43:10.592 couldn’t report — exiting
2 3 45’2025.01.09T08:48:26.748 .Q.dbg
[2] C:\Users\ali\OneDrive – Onyx Commodities\Documents\Projects\kdbq environment\main.q:1: \l ml/ml.q
“””
-
1 Reply
-
Since release 4.0.0 the ML toolkit can use either PyKX or embedPy
https://github.com/KxSystems/ml/releases
Both PyKX and embedPy cannot run in the same process so installing and using only one is best.
To make PyKX available to q processes follow: https://code.kx.com/pykx/3.0/pykx-under-q/intro.html#install
The command to run is:
python -c "import pykx;pykx.install_into_QHOME()"
Log in to reply.