How to get VirtualEnv TensorFlow to work in PyCharm? - python

So I installed tensorflow onto my mac through the main response from here: https://stackoverflow.com/a/33691154/6095482
This means that my tensorflow works in a virtualenv. I can run it through my terminal for scripting purposes, but was wondering if there's any way to port that onto an IDE. I am familiar with pycharm, but couldn't figure out the path routing and virtual environments in order to get it going. Does anyone have any advice on where to take this and which IDE would be good?
Thanks!

You have installed tensorflow in a virtualenv, here's how to get PyCharm to use it:
First you want to add that virtualenv you created and installed tensorflow in, as an interpreter in PyCharm. Please follow this guide from PyCharm makers: Adding Existing Virtual Environment
You will want to select python executable from your virtualenv forlder, e.g. your_venv/bin/python
Once added as an interpreter you want to select it for your project. Here's a guide: Selecting Python interpreter for a project

Related

PyCharm doesn't recognize packages installed via Terminal when environment is used in PyCharm in Pop_OS Linux

I'm using Pop_OS! Linux. Whenever I make a Python Virtual Environment using mkvirtualenv and then pip install packages into it, that environment does not work when I reference it inside pycharm. I set up virtual environments using virtualenv and virtualenvwrapper.
The workaround I've found is to build any virtual environment you need to use inside of PyCharm inside of pycharm, or using the terminal inside of PyCharm. I'm annoyed it doesn't seem compatible with my native terminal, and can't figure out why. I even get slightly different python versions by running python --version both inside the PyCharm terminal and in the native linux terminal, while inside the same virtual environment! It's like Pycharm isn't reading the same information out of the python environment directory as my terminal is.
What is going on here, and how do I harmonize PyCharm with the rest of my linux machine? I'm also new to Pop_OS! Trying it out for the first time is an adventure.
IDK if that will work, but maybe if you install your packages via the "Python packages" tab in PyCharm (down left -> add package next to the research bar that appeared), PyCharm will see them while it installed them itself.

How can I switch virtual environments with Spyder without going to preferences?

I have been programming in Python using Spyder in combination with virtual environments for a while now. Before I started using virtual envs, I installed Spyder using pip. I am not very familiar with paths and environments, but this seemed the way to go when I repetitively ran into package compatibility issues. However, it has been bugging me that every time I want to switch environment, I need to change the working directory to the folder of the environment and I need to change the Python interpreter manually within Spyder.
When I was looking for a solution, I found that it is possible to install Spyder within a virtual environment. I did this and found out that opening Spyder from a virtual environment has the exact same working directory and interpreter as the standalone version (the one I started with). This means that I still have to change the interpreter and working directory manually... Am I doing something wrong or is the only way to switch environments by changing it in the preferences tab within Spyder?
Background info: Using Windows, Spyder 4.2.1, venv as described on https://docs.python.org/3/tutorial/venv.html

Fundamentals question: Django + Python Virtual Env + pyCharm and CLI -- are two separate virtual environments required?

Bottom Line:
I can get everything to work by configuring two separate virtual environments, one for pyCharm and one for the CLI. Is this really necessary or should I be able to use 1 virtual environment for both as I expected?
More Detailed explanation:
I'm very new so this is probably a facepalm type of question so i'll try to be terse.
I'm using Linux Mint, Python 3.6, django 3.0.3, and pyCharm 2019.3.1.
I can create a virtual env using venv in the cli and it works.
I can also create a NEW virtual env in pyCharm through the settings: Project: Interpreter interface, and it works, however it doesn't have venv as an option, it only has virtualenv.
But if I try to activate the virtual env i created in pyCharm from the cli (using virtualenv of course, not venv), it fails hard and thinks i'm using python 2.7 which isn't even installed on my system. If it try to point pyCharm at the virtual env I setup on the cli, I get an error 134.
Is this just a known/expected issue? Must I have two virtual environments for every project I want to access via both pyCharm AND the cli? And I assume this is unrelated but I also find it odd that pyCharm lists my interpreter as python 3.7, which also is not installed on my system. I'm using 3.6 alone.
Thanks for your time.
At this time, I'm going to just answer this as: you need a separate virtual env for each (pyCharm and CLI) as this approach is not difficult or time-consuming and I have not had any issues working in this way.

VS Code error when importing Django module

I'm working on a web app and I use Django as framework.
I'm using VS Code on a macOS.
I get an error when I try to import some Django module. This is a screenshot of my code in error.
The error message is the following:
[pylint] E0401:Unable to import 'django.conf.urls'
I too was facing this error while working with Python virtual environments. In my case, it was happening because I have installed Django on my virtual environment and my base environment didn't contain any module named Django.
Base(Global) environment
and when I use the same command inside my virtual environment (myDjangoEnv)
Fix:
Now what I understood is that pylint uses the python interpreter to analyze our code and mark the error while we write the code.
Since I have three python environments in my system, hence three different python interpreters are available to my VS Code.
So if VS code uses the Python interpreter from the base environment, it will not recognize the Django module (import Error). Therefore, you are required to change the Python interpreter to the one present in your virtual environment.
It sounds complicated but it is pretty simple:
Click on the bottom left of the screen to change python interpreter.
Select from the list of available Python interpreters. Make sure you select the appropriate interpreter with respect to the current project.
Follow steps mentioned in the image. For details, or if that doesn't work, read further!!!
If you can't see your interpreter (installed in the virtual environment) listed in the drop-down list OR selecting interpreters listed don't rectify the error.
Then, you have to give the path of your interpreter (installed in venv) to vs code. Because you might have installed Django only in your venv. Happens when you don't use anaconda to create venv.
Steps to rectify-
To check the path, activate venv and type which python in terminal, this will give path. Copy the path.
Click interpreter on lower left, to pull drop-down, as shown in pic above.
Click enter the interpreter path.
Paste path copied.
This will assign the right interpreter and rectify the error.
Locate your project's virtual environment. In my case, I am working on a Django project and my virtual environment is located on the path below:
C:/Users/abc/Desktop/Virutal36/myLab/Scripts/python.exe
Copy the address of your virtual environment.
On VS Code, Select File > Preferences> Settings to open your User Settings (or use the Ctrl+, shortcut).
Create or modify an entry for python.pythonPath with the full path to your virtual environment and you will be good to go. In my case it is:
C:/Users/abc/Desktop/Virutal36/myLab01/Scripts/python.exe
https://code.visualstudio.com/docs/python/environments#_manually-specify-an-interpreter
Ctrl+Shift+P
Type Python:select interpreter
Now we will get:
Choose Enter interpreter path:
Select Find...
Then
Check correct django version is properly installed and active?
In the active environment, calling this code in python interpreter shouldn't have errors.
from django.conf.urls import url
Check the VS studio python environment
https://code.visualstudio.com/docs/python/environments
You need to select the right environment. So, go to view in tool bar, then select command pallet(ctrl+shift+p), then type "python:select interpreter", then select the right virtual environment where you start you project.
In my case I solved it using the Select Interpreter option from VS Code's Command Palette (Shift + Command + P).
I chose the Python interpreter option which corresponds with the folder in which my virtual environment was and it solved the issue immediately.
Hope it helps :)
Are you using a virtual environment (mkvirtualenv)? In that case you need to make sure you install django and pylint etc., within your virtual environment too, using the following commands.
workon [yourEnvName]
pip install pylint
pip install django
pip install djangorestframework
and so on...for all the modules you want to use.
enter image description here
Choosing a global environment helped me to recover this issue
First check the requirements that you install djangorestframework, second if you work on virtual envs maybe the pylint that you use not check in your virtual env so you can install in local the package.
For me, I opened the nested folder of my Django project. So, Django couldn't find the modules of venv.
Just mentioning, I thought it might be helpful to someone.

How to set default interpreter and keep things in order?

I was required to install anaconda for a CS course and used spyder and Rstudio.
Then, for a different class I used pycharm.
When I type on the command line "python -V" I get:
Python 3.6.1 :: Anaconda 4.4.0 (x86_64)
and I have no idea why it relates the python version I have installed with Anaconda (and why not pycharm?). I understand that the OS runs python 2.7 (shouldn't I get that instead? and when I type python3 -V get which version of python 3 I have?) and when I use something like Pycharm or Spyder I can choose which version I want from the ones I have installed and use it within the program, not for the terminal.
I just want to have everything in order and under control. I don't think I understand what Anaconda really is (to me is like a program that has more programs in it...). How do I keep anaconda to itself ? 1313
Also, should the packages I installed through Terminal work on both pycharm and spyder/anaconda even though when I used pycharm I used python 3.5 and anaconda 3.6?
I think I need definitions and help to get everything in order in my head and the computer.
Pycharm is just an application to help you write code. Pycharm itself does not run python code. This is why in PyCharm, you need to set the interpreter for a project, which could be any python binary. In PyCharm, go to Preferences > Project > Project Interpreter to see where you would set the python environment being used for a given project. This could point to any python installation on your machine, whether that is the python 2.7 located at /usr/bin/python or a virtual environment in your project dir.
The industry standard way to "keep things in order" is to use what are called virtual environments. See here: https://docs.python.org/3/library/venv.html. A virtual environment is literally just a copy of a python environment (binaries and everything) so whatever directory you specify. This allows you to configure your environment to however you need in your project without interfering with other projects you might have. For example, say project A requires django 1.9.2 but project b requires 1.5.3. By having a virtual environment for each project, dependencies won't conflict.
Since you have python3.6, I would recommend going to you project directory in a terminal window. Running python -m venv .venv to create a hidden directory which contains a local python environment of whatever your 3.6 python installation. You could then set your project interpret to use that environment. to connect to it on the command line, run source .venv/bin/activate from where you created your virtual environment. run which python again and see that python is now referencing your virtual environment :)
If you are using a mac (which I believe you are from what you said about python2.7), what likely happened is that your anaconda installer put the Python bin directory on your PATH environment variable. Type in which python to see what the python alias is referencing. You can undo this if you want by editing your ~/.bash_profile file if you really want.
You are more or less correct about anaconda. It is itself another distribution of python and contains a load of common libraries/dependencies that tend to make life easier. For a lot of data analysis, you likely won't even need to install another dependency with pip after downloading anaconda.
I suspect this won't be all too helpful at first as it is a lot to learn, but hopefully this points you in the right direction.

Categories