# Data Dictionary — Tian Mira ADB AA 3604

## CSV Summary (`personnalites-bazi-aa-3604-summary.csv`)

| Column | Type | Nullable | Description | Example |
|--------|------|----------|-------------|---------|
| canonical_profile_id | string | no | Tian Mira 内部标识符 | tmc_adb_... |
| adb_id | string | no | Astro-Databank 标识符 | 12345 |
| name | string | no | 姓名 | profile_example_001 |
| gender | string | yes | 性别 | F |
| birth_date | string | no | 出生日期 (YYYY-MM-DD) | 1970-01-01 |
| birth_time_local | string | yes | 当地出生时间 | 12:00 |
| birth_place | string | yes | 出生地 | Example City |
| birth_country_or_region | string | yes | 国家或地区 | Example Country |
| rodden_rating | string | no | Rodden 等级 | AA |
| year_pillar | string | no | 年柱（干支） | 甲子 |
| month_pillar | string | no | 月柱 | 乙丑 |
| day_pillar | string | no | 日柱 | 丙寅 |
| hour_pillar | string | no | 时柱 | 丁卯 |
| day_master | string | no | 日主（日干） | 丙 |
| source_name | string | no | 来源标识符 | astrodatabank |
| calculation_version | string | no | 计算方法 | advanced_v2 |

## Full JSON / Parquet (`personnalites-bazi-aa-3604.json` / `.parquet`)

The full format contains all CSV columns plus additional technical fields
including all individual stems/branches, element assignments, and internal identifiers.
See the JSON Schema (`schema.json`) for the complete field list.

## NDJSON (`personnalites-bazi-aa-3604.ndjson`)

Same schema as the full JSON, one JSON object per line.
Identical logical content to the full JSON array.

## DuckDB (`tian-mira-adb-aa-3604.duckdb`)

Table `profiles` contains all columns from the Parquet file.
Views: `day_master_statistics`, `element_statistics`, `rodden_statistics`, `country_statistics`.

## Notes

- The CSV is a simplified human-readable view; the Parquet and JSON contain additional structures.
- Empty fields remain empty. No values should be inferred.
- Element percentages (if present) are technical Tian Mira outputs, not astrological judgments.