Environment management with conda and pip in MLOps - Step-by-Step Execution

Process Flow - Environment management with conda and pip

Start: Need new environment

↓

Create conda env

↓

Activate env

↓

Install packages with conda

↓

If package missing

Yes↓

Install package with pip

↓

Use environment

↓

Deactivate env

↓

Delete env if needed

↓

End

This flow shows how to create and manage a Python environment using conda, install packages with conda or pip, then use and clean up the environment.

Execution Sample

MLOps

conda create -n myenv python=3.12
conda activate myenv
conda install numpy
pip install requests
conda deactivate

Create a conda environment, activate it, install numpy with conda, install requests with pip, then deactivate the environment.

Process Table

Step	Command	Action	Result/Output
1	conda create -n myenv python=3.12	Create environment named myenv with Python 3.12	Environment 'myenv' created with Python 3.12
2	conda activate myenv	Activate the environment myenv	Prompt changes to (myenv), environment active
3	conda install numpy	Install numpy package using conda	numpy installed successfully
4	pip install requests	Install requests package using pip inside conda env	requests installed successfully
5	conda deactivate	Deactivate the current environment	Prompt returns to base, environment deactivated
6	conda env list	List all conda environments	Shows 'myenv' and 'base' environments
7	conda remove -n myenv --all	Delete the environment myenv	Environment 'myenv' deleted successfully

💡 Environment deleted or deactivated, no active conda environment

Status Tracker

Variable	Start	After Step 1	After Step 2	After Step 3	After Step 4	After Step 5	After Step 7
Active Environment	base	base	myenv	myenv	myenv	base	base
Installed Packages	base packages	base packages	base packages	base + numpy	base + numpy + requests	base packages	base packages

Key Moments - 3 Insights

Why do we sometimes use pip inside a conda environment instead of conda?

Some packages are not available in conda repositories, so pip is used inside the active conda environment to install those packages, as shown in step 4 of the execution_table.

What happens if you try to install packages without activating the conda environment?

Packages will install in the base environment or system Python, not the intended environment. Step 2 shows activation is needed before installing packages in 'myenv'.

Why do we deactivate the environment after use?

Deactivating returns you to the base environment or system Python, preventing accidental package installs or runs in the wrong environment, as shown in step 5.

Visual Quiz - 3 Questions

Look at the execution_table, what is the active environment after step 3?

A None

B base

C myenv

D root

💡 Hint

Check the 'Active Environment' row in variable_tracker after step 3

At which step does the environment become inactive?

A Step 4

B Step 5

C Step 6

D Step 7

💡 Hint

Look at the 'Active Environment' variable in variable_tracker and the command in execution_table

If you skip step 2 and run 'conda install numpy', where will numpy be installed?

A In the base environment

B In the 'myenv' environment

C Nowhere, command fails

D In a new environment

💡 Hint

Refer to key_moments about activating environment before installing packages

Concept Snapshot

conda create -n envname python=3.12  # create env
conda activate envname                # activate env
conda install package                  # install with conda
pip install package                    # install with pip if missing
conda deactivate                      # leave env
conda remove -n envname --all         # delete env
Use conda envs to isolate projects and avoid conflicts.

Full Transcript

This lesson shows how to manage Python environments using conda and pip. First, you create a new environment with a specific Python version using 'conda create'. Then you activate it with 'conda activate' to work inside it. You install packages preferably with 'conda install', but if a package is missing, you use 'pip install' inside the active environment. After finishing work, you deactivate the environment with 'conda deactivate' to return to the base system. You can list environments with 'conda env list' and delete an environment with 'conda remove -n envname --all'. Activating the environment before installing packages ensures they go to the right place. Deactivating prevents accidental installs in the wrong environment. This process helps keep projects clean and dependencies separate.