Virtual Environments & Python Best Practices

# Create:
python -m venv myenv

# Activate:
source myenv/bin/activate       # Mac/Linux
myenv\Scripts\activate           # Windows

# Install packages:
pip install pandas numpy matplotlib

# Save dependencies:
pip freeze > requirements.txt

# Install from requirements:
pip install -r requirements.txt

A virtual environment is an isolated Python installation with its own packages. Without it, all projects share packages — version conflicts arise when one project needs pandas 1.x and another needs 2.x. Always create a venv per project.

# Good PEP 8 style:
def calculate_growth_rate(current_value, previous_value):
    if previous_value == 0:
        return 0
    return (current_value - previous_value) / previous_value * 100

# Bad style (fails PEP 8):
def calcGrowthRate(c,p):
  if p==0: return 0
  return(c-p)/p*100

PEP 8 is the Python style guide. Key rules: snake_case for functions/variables, 4-space indentation, spaces around operators, lines ≤79 chars, blank lines between functions. Use autopep8 or black to auto-format.

def clean_dataframe(df, drop_cols=None, fill_value=0):
    """
    Clean a DataFrame by dropping specified columns and filling nulls.

    Parameters
    ----------
    df : pd.DataFrame
        Input DataFrame to clean.
    drop_cols : list, optional
        Columns to drop. Default None.
    fill_value : int or float
        Value to fill NaN. Default 0.

    Returns
    -------
    pd.DataFrame
        Cleaned DataFrame.
    """
    if drop_cols:
        df = df.drop(columns=drop_cols)
    return df.fillna(fill_value)

Docstrings document the function for other developers (and your future self). The NumPy/Google style is used in data science projects. Tools like Sphinx auto-generate documentation from docstrings.

import logging

logging.basicConfig(
    level=logging.INFO,
    format="%(asctime)s — %(levelname)s — %(message)s",
    filename="pipeline.log"
)

logging.info("Starting data pipeline")
logging.warning("Missing values found: 152 rows")
logging.error("File not found: sales.csv")

# vs print:
print("done")  # not saved, no timestamp, no severity level

logging writes to files, includes timestamps and severity levels, and can be turned off without changing code. print() only shows in the console. Automated scripts must use logging — without it, errors on scheduled runs are invisible.

import pandas as pd
from typing import Optional, List

def process_sales(
    df: pd.DataFrame,
    region: str,
    top_n: int = 10,
    exclude_cols: Optional[List[str]] = None
) -> pd.DataFrame:
    """
    Process sales DataFrame for a given region.
    """
    result = df[df["region"] == region]
    return result.head(top_n)

Type hints tell other developers (and IDEs) what types a function expects and returns. They are not enforced at runtime but make code far more readable and enable IDE auto-complete and static analysis tools like mypy.