csv-analysis

CSV Analysis

How to analyse CSV data deterministically using tools.

Approach

1. Examine structure first — use read_file to read the first 5-10 lines. Identify columns, data types, and delimiters.

2. Write code, don't guess — always use execute_code with pandas for numerical answers. Never estimate or count manually.

3. Validate results — check row counts, null values, and data types before drawing conclusions.

Template

import pandas as pd

df = pd.read_csv('path/to/file.csv')
print(f"Shape: {df.shape}")
print(f"Columns: {list(df.columns)}")
print(df.describe())

Common pitfalls

• CSV files may have different encodings — try encoding='utf-8' first, then 'latin-1'
• Large files: use nrows=1000 for initial exploration
• Mixed types in columns: pandas may infer wrong dtype, check with df.dtypes