데이터 사양

기록 필드, 파이프라인, 데이터 커버리지에 대한 기술 참조입니다.

Last reviewed: 2026-04-07

Read time 1 min Key sections 4

Who this page helps

기록의 구조와 처리 과정을 자세히 알고 싶은 분

Start with

Next step

tiny wag 소개 Start with the site philosophy and the big picture of the workflow. Open page

Also useful

현재 telemetry 커버리지

현재 기록이 어느 정도 구조화되고 재정리되어 재사용 가능한지 동적으로 요약합니다.

Logged days

1,427 days

Latest diary date 2026-04-09

Refined records

98.9 %

1,412 days include corrected or expanded fields

Structured events

3,713 entries

Split into health events on 1,363 days

Socialization days

74.1 %

1,058 days include a social interaction event

Training logs

810 entries

572 days include structured training logs

파이프라인

  1. 다이어리/아티클 마크다운 생성
  2. 텔레메트리 JSON으로 변환 (convert_diary_to_json_telemetry_v2.py)
  3. 의미 이벤트(행동, 환경, 개입)를 시간 앵커로 추출
  4. 전/후 윈도우(예: -30분~+60분)에서 이벤트를 비교해 연관성을 검토
  5. JSON을 원본 데이터로 저장 (posts/telemetry/ja/telemetry_XXXX.json)
  6. WP-Cron(01:30)으로 SQLite 일일 델타 동기화
  7. 대시보드/FAQ/고정 페이지에서 재사용

필드 참조

모든 공개 필드의 이름, 유형 및 의미를 정리한 참조 테이블입니다.

스키마 버전: 2.2

Basic Info

필드타입Nullable설명
필드age_days타입integerNullable설명Age in days since birth
필드date타입stringNullable설명Record date (YYYY-MM-DD)
필드day_number타입integerNullable설명Days since adoption (1-indexed)
필드weight_kg타입numberNullable설명Body weight (kg)

Behavior & Health Scores

필드타입Nullable설명
필드metrics.activity_level타입integerNullable설명Activity level (1-10, AI-estimated)
필드metrics.appetite타입integerNullable설명Appetite (1-10, AI-estimated)
필드metrics.diet_g타입integerNullable설명Total daily food intake (g)
필드metrics.play_duration_min타입integerNullable설명Play duration (minutes)
필드metrics.socialization_progress타입integerNullable설명Socialization progress (1-10, AI-estimated)
필드metrics.stress_level타입integerNullable설명Stress level (1-10, AI-estimated)

Food Intake

필드타입Nullable설명
필드food.evening_g타입integerNullable설명Evening meal (g)
필드food.morning_g타입integerNullable설명Morning meal (g)
필드food.snack_g타입integerNullable설명Snack/treat amount (g, explicit mentions only)
필드food.total_g타입integerNullable설명Total food intake (g)

Elimination

필드타입Nullable설명
필드elimination.defecation.coprophagy_flag타입booleanNullable설명Coprophagy detected
필드elimination.defecation.failure타입integerNullable설명Defecation accidents
필드elimination.defecation.success타입integerNullable설명Successful defecations
필드elimination.defecation.times타입arrayNullable설명Defecation times (HH:MM format)
필드elimination.urination.failure타입integerNullable설명Urination accidents
필드elimination.urination.success타입integerNullable설명Successful urinations
필드elimination.urination.times타입arrayNullable설명Urination times (HH:MM format)

Health Events

필드타입Nullable설명
필드health_events타입arrayNullable설명Array of health and behavioral events
필드health_events[].description타입stringNullable설명Event description
필드health_events[].severity타입integerNullable설명Severity (0=routine, 1-2=minor, 3+=serious)
필드health_events[].type타입stringNullable설명Event type (vomiting, clinic, grooming, etc.)

Training Logs

필드타입Nullable설명
필드training_logs타입arrayNullable설명Array of training command records
필드training_logs[].command타입stringNullable설명Command name
필드training_logs[].confidence타입numberNullable설명Parser confidence (0.0-1.0)
필드training_logs[].context.arousal타입stringNullable설명Arousal state (calm/excited/mixed/unknown)
필드training_logs[].context.distraction타입stringNullable설명Distraction level (low/high/unknown)
필드training_logs[].context.location타입stringNullable설명Location (home/outside/mixed/unknown)
필드training_logs[].notes타입stringNullable설명Training notes
필드training_logs[].signals타입arrayNullable설명Array of context signals
필드training_logs[].status타입stringNullable설명Mastery status (introduced/in_progress/mastered)

Environment

필드타입Nullable설명
필드environment.delivery_events타입integerNullable설명Number of delivery events
필드environment.separation_duration_min타입integerNullable설명Separation duration (minutes)
필드environment.separation_session_count타입integerNullable설명Number of separation sessions
필드environment.visitor_count타입integerNullable설명Number of visitors
필드environment.weather타입stringNullable설명Weather (sunny/cloudy/rainy/snowy)

Weather Details

필드타입Nullable설명
필드weather_details타입objectNullable설명Detailed weather API data
필드weather_details.cloud_cover_percent타입numberNullable설명Cloud cover (%)
필드weather_details.condition타입stringNullable설명Weather condition (sunny/cloudy/rainy/snowy)
필드weather_details.humidity_percent타입numberNullable설명Humidity (%)
필드weather_details.precip_mm타입numberNullable설명Precipitation (mm)
필드weather_details.precip_prob_percent타입numberNullable설명Precipitation probability (%)
필드weather_details.pressure_hpa타입numberNullable설명Atmospheric pressure (hPa)
필드weather_details.sunrise타입stringNullable설명Sunrise time (HH:MM:SS)
필드weather_details.sunset타입stringNullable설명Sunset time (HH:MM:SS)
필드weather_details.temp_max_c타입numberNullable설명Max temperature (°C)
필드weather_details.temp_min_c타입numberNullable설명Min temperature (°C)
필드weather_details.uv_index타입integerNullable설명UV index
필드weather_details.wind_dir_16타입stringNullable설명Wind direction (16-point compass)
필드weather_details.wind_speed_ms타입numberNullable설명Wind speed (m/s)

Tags

필드타입Nullable설명
필드tags타입arrayNullable설명Array of auto-extracted event tags

시간 단위 인과 추론 (Beta)

시맨틱 매칭으로 일상 맥락과 반응을 검토 가능한 연관성으로 변환합니다.

예: “14:00에 활동 증가”만으로는 원인을 설명할 수 없습니다. 같은 날의 “14:00 배송”, “14:03 짖음”, “14:10 안정” 앵커를 연결해 검토할 체인을 만듭니다: 배송 자극 -> 각성 반응 -> 회복.

  • Output 1: 후보 연관성 (trigger / reaction / recovery)
  • Output 2: 근거 (원문 조각과 시간 차이)
  • Output 3: 신뢰도 (high / medium / low)
  • Output 4: 다음 수집 지시 (무엇을 추가 측정할지)

이 연관성은 예측에서 다시 확인하고 트레이닝 가이드에서 실행으로 연결하세요.

다음 읽기