Loading and Inspecting Data

First steps with any dataset

import pandas as pd
 
df = pd.read_csv("dataset.csv")
 
# Always run these first
print(df.shape)           # how big is it?
print(df.head())          # what does it look like?
print(df.info())          # column types, missing values
print(df.describe())      # statistics for numeric columns
print(df.isnull().sum())  # missing values per column

Common formats

Format	Load with	Notes
CSV	pd.read_csv()	Universal, but slow for large files
Parquet	pd.read_parquet()	Fast, compressed, preserves types — use for large data
JSON	pd.read_json()	Nested data, API responses
Excel	pd.read_excel()	Needs openpyxl
SQL	pd.read_sql()	Direct from database

Questions to ask about any dataset

• How many samples? How many features?
• What are the column types? (numeric, categorical, text, datetime)
• How much is missing? Is it random or systematic?
• What is the target variable? Is it balanced?
• Are there duplicates?

Links

• Pandas Essentials
• Data Cleaning
• Exploratory Data Analysis