Questions on the SI language - page 4

 
Maxim Dmitrievsky:

I dare say that C is not the best choice. It is certainly a well-respected and respected language, but at the moment it is very, very narrow in scope. Maybe while you're not too far into it, you should change your choice. In favour of C++ or the same Sharp?

 
Andrei Novichkov:

I dare say C doesn't seem like the best choice. It's a respectable and well-respected language, but at the moment it's very, very limited in scope. Maybe while you're not too far into it, you should change your choice. In favour of C++ or the same Sharp?

It says in the beginning that I need to rewrite the library from C to µl, no other purpose. Many interesting libraries for python (all of them) are written in C. So it is necessary to navigate in it.

i.e. i've done some research in python, and i don't need it as a gasket anymore, i can rewrite everything i need in µl, if my hands allow it

plus I want to understand the logic of the lib.

I'm already halfway there, it's not that complicated.
 
Igor Makanu:

viapip freeze > requirements.txt did a one-to-one package installation on the laptop as on the PC, but alas, no miracle, still an error

here attached the check and error log, error googled, solution... And a million solutions , on PC installed on laptop... probably will now start copying python libs and kill python, then from scratch... (((

you not installing the gpu version? they say that the error occurs if the card does not support or CUDA is not installed

check if you need 64 bit python

it makes sense to try a pip3 install, not a pip, if other versions of python have been installed before

 
Igor Makanu:

I have no graphics card on my PC, everything works there, I am trying to install on my laptop by analogy without GPU, it does not work either, and to install GPU you need NVidia CUDO to install.... i think it will be a big lottery, yesterday i looked up WHERE the installer went. it's a lot of ticks ))))

Pythons are the same 3.7 - 64 bit, i think i had some g++ compiler on my PC but i can only see the software i have on my PC tomorrow

ok, not critical but it's a shame ((((

the gpu is not needed for small projects and will always lose out to the cpu in terms of speed

well then something with the mb compiler, no other options

g++ is a minGW compiler command
 
Igor Makanu:

hmm, I think it's the condo, I installed Spyder on my laptop with the condo because Picharm lags, it's probably the condo's fault, but the condo has minGW, maybe the paths are lost somewhere

must not be destined to work with highly intellectual TensorFlow, I will have to use examples under Scikit-learn

konda is evil, do not use )) maybe the compiler has not registered in the PATH... but then it would give you a recommendation, but not for sure for konda

you can google colab in the browser, there is already a tensor there :) and examples are even

 
Maxim Dmitrievsky:

conda evil, don't use it ))

you can google colab in your browser, there's already a tensor there :) and there are even examples

Thanks, but that's enough for now, I'll do something else

here's a pretty interesting article, really liked the example with a coin, unlike the discussion here ))))

https://habr.com/ru/company/ods/blog/322716/

PyMC3 — MCMC и не только
PyMC3 — MCMC и не только
  • habr.com
Привет, Хабрахабр! В этом посте уже упоминался PyMC3. Там можно почитать про основы MCMC-сэмплирования. Здесь я расскажу про вариационный вывод (ADVI), про то, зачем все это нужно и покажу на довольно простых примерах из галереи PyMC3, чем это может быть полезно. Одним из таких примеров будет байесовская нейронная сеть для задачи классификации...
 
Igor Makanu:

Thanks, but that's probably enough for now, I'll get on to other things

Here's an article, pretty interesting, really liked the example with the coin, not unlike the discussion here ))))

https://habr.com/ru/company/ods/blog/322716/

Nice album, used it. Will read the article later.

by the way, spyder is placed on the bare python, after a couple of manipulations

the easiest way is to put jupyter notebook or jupyter lab, also on bare python
 
Maxim Dmitrievsky:

It's a great library, I've been using it. I'll read the article later.

by the way, spyder installs on bare python after a couple of manipulations

the easiest way is to use jupyter notebook or jupyter lab, also on bare python

I'll probably clean out my laptop soon, kill everything, then reinstall.

Spyder already do not use, like a Wing IDE began to get used to, seems satisfied, but Picharm certainly cooler, but my laptop is loaded very long -memory 4Gb not enough apparently him

 
Igor Makanu:

I'll probably clean out my laptop soon, kill everything, then reinstall

Spyder no longer use, somehow to Wing IDE began to get used to, it seems to suit, but Picharm certainly cooler, but the laptop loads very long - memory 4Gb not enough apparently him

It eats RAM and drains battery. Jupiter laptop is the most forgiving

Many people even use qt console
 
Maxim Dmitrievsky:

It says in the beginning that I need to rewrite a library from C to µl, no other purpose. Many interesting python libraries (all of them) are written in C....

In C exactly? I didn't know that. Interesting, thanks for the info.

Reason: