PyQ question

Community Support
1

  1. Doenloaded PyQ source from 
    https://pypi.python.org/pypi/pyq/3.7.1

  2. Exported QHOME env variable (OS is 64bit Linux 2.6.18-400.1.1.el5)

  3. Intalled using 

python3 setup.py intall

  1. I can’t find where the shared libraries are placed. 

Specifically -> py.so. p.so

When I fire up Q loading python.q, I get the below error,

$> q python.q

k){0N!x y}

'/home/testuser/q/l32/py.so: cannot open shared object file: No such file or directory

@

“q”

“py:PYSO 2:(`py;3)”

  1. What am I missing? 

Thanks!

2
  1. I checked the build logs and there is no entry of py.so or p.so being built.
  2. The only entry of .so being built is this (from build logs)

    _k3.cpython-34m.so

Tks!

3

On Thursday, August 11, 2016 at 3:50:44 PM UTC-4, merc7791 wrote:

  1. Intalled using 

python3 setup.py intall

Why don’t you use the recommended command?

pipinstall-ihttps://pyq.enlnt.com--no-binarypyqpyq

If you want a specific version, do

pipinstall-ihttps://pyq.enlnt.com--no-binarypyqpyq==3.7.1

 If you want to install the already downloaded source,

pipinstall **/path/to/source**

should work as well.

4

How did you solve your problem then ? It tells me the same thing

Le jeudi 11 août 2016 22:49:20 UTC+2, merc7791 a écrit :

  1. I checked the build logs and there is no entry of py.so or p.so being built.
  2. The only entry of .so being built is this (from build logs)

    _k3.cpython-34m.so

Tks!

5

I found out that 64-bit installation requires a paid license. 

I used qPython instead.

Thanks!

6

On Tuesday, September 20, 2016 at 4:54:13 AM UTC-4, merc7791 wrote:

I used qPython instead.

qPython is a fine library and may suit your needs.  Note, however, that PyQ and qPython take two fundamentally different approaches to working with kdb+ data in Python.  qPython is a traditional database communications library that lets you connect to a kdb+ instance, get the serialized data over an IPC connection, convert it to Python objects that can be manipulated in Python, serialized and sent back to kdb+ for storage.  PyQ is different.  Instead of bringing kdb+ data to Python it brings Python to kdb+.  This eliminates IPC, serialization and copying overhead and may increase the performance of some applications 10-100 times.  See <https://pyq.enlnt.com/slides\> for more details.