DST Table Info Skill

DST Table Info Skill

Purpose

Get complete metadata and structure for a DST table. This is an essential step before fetching data - it tells you what dimensions are available, what filters you can apply, and what to expect in the data.

When to Use

• Before fetching data to understand table structure
• To see available variables and dimensions
• To determine appropriate filters for data fetch
• To understand table updates and units of measurement
• User asks "what's in this table?"
• After finding a table ID with dst-tables skill

How to Use

Basic Usage (JSON output)

Get table metadata:

python scripts/​api/​get_tableinfo.py --table-id <TABLE_ID>

Verbose Format (Human-readable)

Get formatted, easy-to-read output:

python scripts/​api/​get_tableinfo.py --table-id <TABLE_ID> --verbose

Save to File

Save metadata for reference:

python scripts/​api/​get_tableinfo.py --table-id <TABLE_ID> --output info.json

Expected Output

The script returns complete table metadata including:

Basic Information

• id: Table identifier
• text: Table name
• description: Detailed description
• unit: Unit of measurement
• updated: Last update timestamp

Variables/​Dimensions

• id: Variable identifier (e.g., "OMRÅDE", "TID")
• text: Variable description
• values: All available values for that variable
• time: Whether it's a time dimension

Example output structure:

{
  "id": "FOLK1A",
  "text": "Population at the first day of the quarter",
  "description": "Population by region, age, and sex",
  "unit": "number",
  "updated": "2025-10-15T10:30:00",
  "variables": [
    {
      "id": "OMRÅDE",
      "text": "Region",
      "values": [
        {"id": "000", "text": "Whole country"},
        {"id": "101", "text": "Copenhagen"}
      ]
    },
    {
      "id": "TID",
      "text": "Time",
      "time": true,
      "values": [
        {"id": "2020Q1", "text": "2020Q1"},
        {"id": "2020Q2", "text": "2020Q2"}
      ]
    }
  ]
}

Key Information to Extract

When analyzing table info, pay attention to:

1. Table Description: What data does it contain?
2. Unit: What do the numbers represent? (persons, kroner, percentage, etc.)
3. Variables: What dimensions can you filter by?
4. Available Values: What options exist for each variable?
5. Time Coverage: What time periods are available?
6. Last Updated: How fresh is the data?
7. Elimination Flag: Whether variables can be aggregated (see below)

Understanding Variable Properties

Each variable has important metadata:

elimination flag:

• true: Variable can theoretically be aggregated/​eliminated
• false: Variable must be explicitly specified
• Note: BULK format requires ALL variables regardless of this flag

time flag:

• true: This is the time dimension
• Time values have special formats: 2024 (annual), 2024K1 (quarterly), 2024M01 (monthly), 2024U01 (weekly)
• Time variables typically have elimination: false

map property:

• Present if variable has geographic mapping
• Example: "map": "Denmark_municipality_07"
• Indicates visualization possibilities

Variable Value Codes

Important patterns in value IDs:

• Aggregate codes: "TOT", "IALT", "000" (whole country) - represent totals
• Time codes: "2024K1" = 2024 Quarter 1 (K=Kvartal), "2024M01" = 2024 January
• Geographic codes: "000" = Whole Denmark, "101" = Copenhagen, etc.
• Use exact codes from tableinfo in data requests

Next Steps

After understanding the table structure:

1. Decide on filters (optional): Based on variables and values
2. Fetch the data using dst-data skill:

   python scripts/​fetch_and_store.py --table-id <TABLE_ID>
   

3. With filters (if needed):

   python scripts/​fetch_and_store.py --table-id <TABLE_ID> --filters '{"OMRÅDE":["000"]}'
   

Examples

Example 1: Get table info for FOLK1A

python scripts/​api/​get_tableinfo.py --table-id FOLK1A

Example 2: Verbose, readable output

python scripts/​api/​get_tableinfo.py --table-id FOLK1A --verbose

Example 3: Save to file

python scripts/​api/​get_tableinfo.py --table-id FOLK1A --output folk1a_structure.json

Example 4: Check multiple tables

# Get info for population table
python scripts/​api/​get_tableinfo.py --table-id FOLK1A --verbose

# Get info for employment table
python scripts/​api/​get_tableinfo.py --table-id AUP01 --verbose

Tips

Before Fetching Data

• Always check tableinfo first: Large tables can take time to download
• Calculate expected cells: Multiply value counts to estimate size
• Note available filters: Limit data size by filtering dimensions
• Check date range: Understand temporal coverage (first/​last periods)
• Identify ALL variable IDs: BULK format requires all specified

Reading Metadata

• Understand variables: Know what you can group/​filter by
• Check elimination flags: But remember BULK needs all variables anyway
• Note time formats: Quarterly (K), Monthly (M), Weekly (U), Annual
• Find aggregate codes: TOT, IALT, 000 are usually totals
• Use verbose mode: Easier to read when exploring interactively
• Save JSON: Keep machine-readable format for programmatic use

Performance Considerations

• Caching: Tableinfo responses are cacheable (data structure rarely changes)
• Case-insensitive: Table IDs work in any case ("FOLK1A" = "folk1a")
• Response time: Typically 200-400ms
• Call once: Metadata doesn't change frequently, cache locally if needed

Common Variables (Reference)

Typical DST variable names:

• OMRÅDE: Geographic region
• TID: Time period
• ALDER: Age
• KØN: Sex/​Gender
• BRANCHE: Industry
• UDDANNELSE: Education level
• INDKOMST: Income bracket

Note: Variable names are usually in Danish.

Understanding Values

Each variable has a set of possible values:

• Hierarchical: Some variables have parent-child relationships
• Time values: Format varies by table:
  • Annual: "2024"
  • Quarterly: "2024K1" (K=Kvartal/​Quarter)
  • Monthly: "2024M01"
  • Weekly: "2024U01" (U=Uge/​Week)
• Region codes: Numeric codes like "000" (whole country), "101" (Copenhagen)
• Categories: Text or numeric codes for classifications
• Aggregate values: "TOT", "IALT", "000" typically represent totals

Analyzing Metadata for Data Requests

Step 1: Count total cells

cells = var1_values × var2_values × ... × varN_values

If cells > 1,000,000: Must use BULK format

Step 2: Identify required variables

• BULK format: ALL variables must be specified
• Check variable IDs from "variables" array
• Note which are time dimensions ("time": true)

Step 3: Select appropriate values

• Use "*" for all values
• Use specific IDs for filtering
• Use patterns for partial selection: "2024*", "*K1", etc.
• Use nth-rules for time: "(1)" = latest, "(-n+5)" = last 5

Troubleshooting

"EXTRACT-NOTFOUND: Table not found"

Error message: "Tabellen blev ikke fundet"

• Verify table ID is correct (though case-insensitive)
• Check table exists using dst-tables skill first
• Try with different case if needed
• Verify table is active (not discontinued)

Empty or Limited Values

• Some tables may have limited options for certain variables
• Check if table is being updated (look at "updated" field)
• Inactive tables may have outdated metadata

Large Output

• Use --verbose for human-readable summary
• Use --output to save full JSON for later analysis
• Large tables have many variables/​values - normal behavior

Network Errors

• Check internet connection
• DST API may be temporarily unavailable
• Typical response time: 200-400ms
• Retry after a few seconds if timeout occurs

Understanding Complex Structures

• Variables with many values (100+) are common
• Time variables often have 50+ periods
• Geographic variables can have 100+ regions
• This is expected - use filters when fetching data