1. DataFrame: Basic DataFrame Properties

Understand your DataFrame’s structure at a glance.

• df.index → Returns the row index labels of the DataFrame.
• df.columns → Returns the column labels as an Index.
• df.dtypes → Returns a Series with each column’s data type.
• df.axes → Returns a list with the row axis and column axis.
• df.shape → Returns a tuple (rows, columns) representing dimensions.
• df.ndim → Returns the number of dimensions (always 2 for DataFrame).
• df.size → Returns total number of elements (rows × columns).
• df.memory_usage() → Shows memory usage for each column plus index.
• df.empty → Returns True if DataFrame has no elements.
• df.attrs → Dictionary for storing custom user metadata.
• df.info() → Prints concise summary: index dtype, columns, non-null counts.

1. DataFrame Properties Reference

Professional Workflow Guide

Data Inspection Protocol

1. Initial Check:

python

Copy

df.info()  # Verify structure and nulls
print(df.shape)  # Confirm expected size

2. Memory Optimization:
Copy

# Convert appropriate columns to category
cat_cols = df.select_dtypes(include='object').nunique()

for col in cat_cols[cat_cols < 100].index:
    df[col] = df[col].astype('category')

3. Type Validation:
Copy

expected_dtypes = {
    'Date': 'datetime64[ns]',
    'Price': 'float64'
}
for col, dtype in expected_dtypes.items():
    assert df[col].dtype == dtype, f"{col} has wrong dtype"

4. Production-Grade Checks:
Copy

def validate_dataframe(df):
    assert not df.empty, "Empty DataFrame"
    assert df.index.is_unique, "Non-unique index"
    assert df.columns.is_unique, "Duplicate columns"
    return True

Performance Considerations

Best Practice: For large DataFrames, avoid repeated calls to memory_usage(deep=True) in production code.