Skip to content
datastudyguides

DP-750 · 3.3.2 · Prepare and process data

Choose column types

Choose appropriate column data types

Use the narrowest type that preserves required values, semantics, precision, and downstream operations.

Inspect types in the notebook output

Azure Databricks notebook output table showing column data types and null values in a DataFrame.

Databricks surfaces each column's type directly beside the data. That makes incorrect typing easy to spot before it propagates; the diagrams below focus on choosing the most appropriate type.

Let meaning choose the type

  • Exact value with fixed precisionDECIMAL
  • Identifier whose characters matterSTRING
  • Business date or event timeDATE or TIMESTAMP
  • Whole-number measureInteger type with sufficient range
Column type

A type should express what the value means, not only how it arrived from the source.

Preserve semantics

  • Better choice

    Currency
    DECIMAL
    Postal code
    STRING
    Timestamp
    TIMESTAMP

  • Risky choice

    Currency
    Floating point
    Postal code
    Integer
    Timestamp
    Raw STRING
Common mistakes

Exact financial values need fixed precision. Identifiers should stay strings when arithmetic has no business meaning.

Avoid string-everything models

Typed columns make invalid values easier to detect and downstream calculations easier to reason about.

Key points

  • Numeric: Match range and precision requirements
  • DECIMAL: Prefer fixed precision for exact financial values
  • Date and timestamp: Preserve temporal semantics
  • String: Use for identifiers when arithmetic is not meaningful

Exam trap

Do not store every field as STRING just because raw ingestion arrives as text.

Sources · 3 Microsoft Learn pages Show ▾

Checked on October 4, 2026. Report a problem

Go deeper on Microsoft Learn

Checked against Microsoft Learn on October 4, 2026. Report a problem with this card

How well do you know this?