Data Format Reference

This section contains detailed reference documentation for OpenPlanetData datasets formats.

Data Format Reference

JSON Format

All JSON datasets follow consistent formatting:

{
  "version": "1.0.0",
  "generated_at": "2024-01-15T00:00:00Z",
  "license": "CC-BY-4.0",
  "data": [
    // Array of records
  ]
}

CSV Format

CSV files include a header row with field names matching the JSON field names.

alpha2,alpha3,name,capital,region
US,USA,United States,Washington D.C.,Americas
FR,FRA,France,Paris,Europe

Parquet Format

Parquet files use the same schema as JSON with appropriate type mappings:

JSON Type	Parquet Type
string	UTF8
number (int)	INT64
number (float)	DOUBLE
boolean	BOOLEAN
array	LIST
object	STRUCT

Version Numbering

Datasets follow semantic versioning:

• MAJOR - Breaking changes to schema or data format
• MINOR - New fields or data additions
• PATCH - Bug fixes and corrections

Checksums

All releases include SHA256 checksums:

# Verify download integrity
sha256sum -c checksums.txt

Rate Limits

When accessing releases via GitHub:

• Unauthenticated: 60 requests/hour
• Authenticated: 5000 requests/hour

For high-volume access, download datasets and host locally.