Skip to content

Attribute Tables

Goal: Learn to read, edit, query, and calculate attribute table fields — the data that makes a map useful.

What you'll learn

  • Anatomy of an attribute table
  • Field types and when to use each
  • Field Calculator basics
  • How attribute tables connect to spatial data

Anatomy of an attribute table

Every vector layer has an attribute table. Each row = one feature. Each column = one field (attribute).

ObjectID Name Population Area_km2 Type
1 Atlanta 498,715 351 City
2 Athens 127,315 308 City
3 Augusta 202,081 791 City
  • ObjectID — auto-generated unique ID. Don't edit it.
  • Geometry / Shape — hidden column with the feature's geometry.
  • The rest — your data.

Field types

Picking the right field type matters for storage, performance, and querying.

Type What it stores Example
Short integer Whole numbers, ±32,767 Census tract count
Long integer Whole numbers, ~±2 billion Population
Float / Double Decimal numbers Latitude, area, NDVI
Text (String) Letters, numbers, symbols City name, ZIP code
Date Dates and times Survey date
Blob / Raster Binary Photos, attachments

ZIP codes and IDs

Store ZIP codes, FIPS codes, and phone numbers as text, not numbers. Otherwise leading zeros disappear (07083 → 7083).

Calculating fields

Use the Calculate Field tool to populate values automatically.

# Calculate area in square km
!shape.area@SQUAREKILOMETERS!

!FIRST_NAME! + " " + !LAST_NAME!


def classify(pop):
    if pop > 100000:
        return "Urban"
    elif pop > 10000:
        return "Suburban"
    else:
        return "Rural"
Then in the calculator: classify(!POPULATION!)

!POPULATION! / !AREA_KM2!

Selecting and filtering

Two ways to filter:

  1. Select by Attributes — WHERE STATE_NAME = 'Georgia' AND POP > 100000 — see SQL for GIS.
  2. Definition Query — temporarily hides features that don't match a condition. Layer behaves as if the others don't exist.

Definition Query vs Selection

  • Selection = highlights matching features (still visible).
  • Definition Query = filters them out of the layer entirely (only matches show / process).

Joining attribute tables

You can attach a table (CSV, table, or another layer's attributes) to a feature class using a shared field.

flowchart LR
    L[Counties layer] -- FIPS = FIPS --> J((Join))
    T[Population CSV] -- FIPS --> J
    J --> R[Counties + population]

    classDef l fill:#dbeafe,stroke:#1e40af,color:#1e3a8a
    classDef t fill:#fef3c7,stroke:#f59e0b,color:#92400e
    classDef r fill:#dcfce7,stroke:#10b981,color:#065f46
    class L l
    class T t
    class R r
    classDef j fill:#fff,stroke:#4338ca,color:#312e81
    class J j

→ Detailed walkthrough: Joins & Relates

Domains and subtypes (advanced)

In a file geodatabase, you can define:

  • Domain — list of allowed values for a field (e.g., LandUse can only be Residential / Commercial / Industrial)
  • Subtype — different rules for different categories of feature

These reduce data entry errors and enforce consistency.

Best practices

Clean attribute tables = clean analysis

  • Use descriptive field names (POP_2020, not F1).
  • Don't store calculated values that you can derive (e.g., density). Calculate when needed.
  • Document fields with aliases (Field Properties → Alias).
  • Validate types — a numeric field with text in it will break analysis.

Practice

Try this

  1. Download US Counties shapefile (TIGER/Line).
  2. Add a CSV with FIPS, MEDIAN_INCOME for each county.
  3. Join the CSV to the counties using the FIPS field.
  4. Calculate a new field: INCOME_BIN = Low / Mid / High based on MEDIAN_INCOME.
  5. Use Definition Query to show only INCOME_BIN = 'Low' counties.

Next up

→ Spatial Joins — joining tables based on location.