Exploratory Data Analysis (EDA) Workflow

The EDA Checklist

1. **Shape**: `df.shape` — How many rows and columns?

2. **Types**: `df.dtypes` — What data types are there?

3. **Nulls**: `df.isnull().sum()` — Where is the missing data?

4. **Statistics**: `df.describe()` — Summary stats.

5. **Distributions**: Histograms for numerical columns.

6. **Correlations**: Heatmap.

7. **Outliers**: Box plots.

Pro Tip

Spend 20% of your project time on EDA. It prevents 80% of downstream errors.

*Day 120: Handling Outliers.*