Why is my flask app running with no modules installed in my environment?

Question

I am using Anaconda to build environments. I've been using pip install my_module the entire time, and I've recently realized this is incorrect. The environment I made is called dp_offfsets_environment and I have been able to run my flask app I wrote in IntelliJ using an SDK linked to this environment.

Why is this? Seen below, when calling conda list no packages are installed under my environment so I would expect my flask app to fail.

What is pip list displaying below? Is it showing packages installed in my base environment?

How do you recommend uninstalling these packages and moving them into the actual environment?

(dp_offsets_environment) C:\WINDOWS\system32>pip list
Package                Version
---------------------- -------
click                  8.0.3
colorama               0.4.4
cycler                 0.10.0
Flask                  2.0.2
fonttools              4.29.1
importlib-metadata     4.10.1
itsdangerous           2.0.1
Jinja2                 3.0.3
kiwisolver             1.3.1
MarkupSafe             2.0.1
matlab                 0.1
mysql-connector-python 8.0.26
packaging              21.3
pandas                 1.3.0
Pillow                 8.3.1
pip                    22.0.2
pyparsing              2.4.7
python-dateutil        2.8.1
pytz                   2021.1
setuptools             47.1.0
six                    1.16.0
typing_extensions      4.0.1
Werkzeug               2.0.2
wheel                  0.37.1
zipp                   3.7.0

(dp_offsets_environment) C:\WINDOWS\system32>conda list
# packages in environment at C:\Users\ckurtz\.conda\envs\dp_offsets_environment:
#
# Name                    Version                   Build  Channel

Answer 1

When you run conda create -n dp_offsets_environment, then conda creates an empty environment, i.e. with no packages. conda does not install python, pip or anything else in the environment. So when you run pip with that env activated, you will still get whatever pip is first located in your PATH (most likely from your base env).

So whenever you ran pip install nothing was installed to your currently active conda env and your flask app was probably not using that conda env, but also used the python that corresponds to your pip.

You need to install python/pip to your env and then install to your env.

See this as an example (I added annotation with ###):

(base) C:\Users\FlyingTeller>where pip
C:\Users\FlyingTeller\miniconda3\Scripts\pip.exe ### We are in base, so pip from base is found

(base) C:\Users\FlyingTeller>conda create -n someEnv
Collecting package metadata (current_repodata.json): done
Solving environment: done


Please update conda by running

    $ conda update -n base conda



## Package Plan ##

  environment location: C:\Users\FlyingTeller\miniconda3\envs\someEnv

### Note: No packages are being installed

Proceed ([y]/n)? y

Preparing transaction: done
Verifying transaction: done
Executing transaction: done
#
# To activate this environment, use
#
#     $ conda activate someEnv
#
# To deactivate an active environment, use
#
#     $ conda deactivate


(base) C:\Users\FlyingTeller>conda activate someEnv

(someEnv) C:\Users\FlyingTeller>where pip
C:\Users\FlyingTeller\miniconda3\Scripts\pip.exe ### <----- pip still from base env

(someEnv) C:\Users\FlyingTeller>conda install python pip
Collecting package metadata (current_repodata.json): done
Solving environment: done


Please update conda by running

    $ conda update -n base conda



## Package Plan ##

  environment location: C:\Users\FlyingTeller\miniconda3\envs\someEnv

  added / updated specs:
    - pip
    - python


### Download and installation pruned
Preparing transaction: done
Verifying transaction: done
Executing transaction: done

(someEnv) C:\Users\FlyingTeller>where pip
C:\Users\FlyingTeller\miniconda3\envs\someEnv\Scripts\pip.exe ### Only now pip is in this env
C:\Users\FlyingTeller\miniconda3\Scripts\pip.exe

(someEnv) C:\Users\FlyingTeller>

---

Side Note

I've been using pip install my_module the entire time, and I've recently realized this is incorrect

This is a misconception of a commonly given advice "Do not use pip in a conda env". This advice is always a bit strong. There is no reason to refrain from using pip in a conda env all together. There is a good reason to be careful and aware of certain things that can easily break your env. There is also a good post on anaconda.org about this.

Answer 2

For anyone looking back on this, I discovered I had another instance of Anaconda on my machine. My company gives out laptops with Spyder installed which by default, has it's own version of Anaconda. I did not know Spyder came with Anaconda so I accidently installed two versions of Anaconda during my initial setup. To fix this I had to go through and uninstall both instances of Anaconda on my machine including their packages. Once completely removed, I reinstalled Anaconda, and my problems were alleviated.