Reading and writing CSV files

Two ways: csv module and pandas

The csv module is built into Python. pandas is the third-party library we'll lean on heavily. For 95% of work, pandas is what you want. For tiny scripts or constrained environments, the csv module is fine.

The pandas one-liner

import pandas as pd
df = pd.read_csv("sales.csv")
print(df.head())

Done. You have a DataFrame.

Writing a CSV

df.to_csv("clean.csv", index=False)

index=False suppresses pandas' row-number column, which is usually what you want.

Useful read_csv options

pd.read_csv("sales.csv",
            sep=",",                  # delimiter
            encoding="utf-8",         # "latin-1" for some legacy exports
            skiprows=2,               # skip first two rows
            header=0,                 # row to use as headers
            usecols=["date","amount"],# read only these columns
            parse_dates=["date"],     # turn date column into real dates
            dtype={"id": str},        # force this column to be string
            nrows=1000,               # read only first 1000 rows
           )

The csv module (for when you can't use pandas)

import csv

with open("sales.csv") as f:
    reader = csv.DictReader(f)
    rows = list(reader)

print(rows[0])      # {'date': '2026-01-05', 'amount': '120', ...}

Every value comes back as a string. You convert as you go.

Common CSV headaches and fixes

Walkthrough: end-to-end CSV cleanup

import pandas as pd
df = pd.read_csv("sales_raw.csv", parse_dates=["date"])

df.head()
df.info()
df.describe()

df.columns = df.columns.str.strip().str.lower()
df = df.dropna(subset=["customer"])
df["amount"] = df["amount"].astype(float)

df.to_csv("sales_clean.csv", index=False)