DATASET · research

Daybreak 盘前观察清单

research.daybreak_premarket_watchlist

研究成品

这是已经产出的研究信号与市场状态中的“Daybreak 盘前观察清单”数据集；一行由 trade_date、symbol 定位，主要字段包括 instrument_id、rank、priority_score、trigger_day_count、detector_count。

下载本类 JSON 下载完整 JSON.gz 先获取 Key

字段27

关键字段2

更新口径producer_schedule

时间列trade_date

单页上限5000

分页Cursor

PIT不支持

规模1K-9K

WHAT IT ANSWERS

这张表适合回答什么

读取已完成的研究信号和市场状态
复盘信号随行情变化的结果

GRAIN & TIME

一行代表什么

业务粒度: 每行代表“Daybreak 盘前观察清单”的一条记录；逻辑粒度由 trade_date × symbol 共同确定。
时间口径: 以 trade_date 为主要查询时间轴；除下方限制特别说明外，时间戳按 UTC、日期按来源市场日历解释。
更新节奏: 由对应生产任务或事件触发更新，不承诺固定日历频率；查询前先检查最新时间。

KEY FIELDS

先看这些字段

trade_date: 对应市场日历中的交易日。
symbol: 证券或交易对的展示代码。

RECOMMENDED JOINS

怎样和其他表关联

ref.instrumenton instrument_id

用 instrument_id 补充证券代码、名称、市场、交易所和资产类型。

ref.calendaron market + trade_date（有 market 时）

需要计算交易日偏移时联接交易日历；无 market 字段时先确定适用市场。

HOW TO QUERY SAFELY

查询方法与已知限制

先用 catalog 确认当前契约；可按 catalog 中列出的 filters 进一步缩小范围；单页最多 5000 行，响应有 has_more 时原样传回 next_cursor 继续分页。

Score correction notice: priority_score now returns the real priority score of each watchlist name on every date (the list is ranked by it; higher ranks first). For trade dates from 2026-08-20 until this notice appeared, the API returned a rank placeholder in priority_score instead of the score: 3.0 for rank 1, falling in equal steps to 2.0 for rank 30. It only restated rank and carried no score information. If you downloaded rows dated 2026-08-20 or later before this notice, download them again. The placeholder is still available, under its own name, in the new column rank_display_score. Rows dated before 2026-08-20 did not change. 分数更正说明：priority_score 现在在全部日期上都返回每只股票真实的优先级分数（名单按它排序，越高越靠前）。交易日 2026-08-20 起、到本说明出现之前，接口在 priority_score 里返回的其实是按名次换算的占位值（第 1 名 3.0，按等差递减到第 30 名 2.0），只是名次的另一种写法，不含分数信息。如果你在本说明出现前下载过 2026-08-20 及之后的数据，请重新下载。该占位值仍以新列 rank_display_score 提供。2026-08-20 之前的数据没有变化。

MACHINE CONTRACT

筛选、排序与分页

UNVERSIONED

默认排序: trade_date, symbol
唯一键: trade_date, symbol
读取模式: canonical_final_active
最大时间窗: 未单独限制

可用 filters

instrument_id→ instrument_id · integer

symbol→ symbol · string

trade_date→ trade_date · string

POINT IN TIME

可复现读取

NOT AVAILABLE

该资产没有 PIT 契约。普通 rows 端点仍按当前公开读模式返回数据；不要把它表述成历史时点快照。

COPY-READY CALLS

怎么调用

同一数据集可用 HTTP、Python SDK、CLI 或 MCP；Key 始终只通过环境变量/Bearer Header 传递。

READ ONLY

HTTP

curl --fail-with-body -H "Authorization: Bearer $FRIENDS_DATA_API_KEY" "https://api.alphahubs.uk/api/v1/data/datasets/research/daybreak_premarket_watchlist/rows?limit=100&instrument_id=1"

PYTHON SDK

from friends_data_api import FriendsDataClient, Settings

async with FriendsDataClient(Settings.from_env()) as client:
    page = await client.get_dataset_rows('research', 'daybreak_premarket_watchlist', limit=100, filters={'instrument_id': 1})
    print(page.data)
    if page.has_more:
        print("next cursor:", page.next_cursor)

CLI

friends-data get-dataset-rows research daybreak_premarket_watchlist --limit 100 --filter instrument_id=1

MCP TOOL CALL

{
  "arguments": {
    "dataset": "daybreak_premarket_watchlist",
    "filters": {
      "instrument_id": 1
    },
    "limit": 100,
    "namespace": "research"
  },
  "tool": "get_dataset_rows"
}

FIELD DICTIONARY

27 个返回字段

descriptionSource 可在 JSON 中区分仓库注释、标准术语表和名称推导说明。

筛选字段

#	字段	ClickHouse 类型	JSON	可空	角色	说明
1	`trade_date`date	`Date`	`string / date`	no	time, filter, default_order, warehouse_primary_key, contract_unique_key	Trading-session date in the dataset's market calendar.标准术语表
2	`symbol`	`LowCardinality(String)`	`string`	no	filter, default_order, warehouse_primary_key, contract_unique_key	Display or trading symbol for the instrument.标准术语表
3	`instrument_id`	`UInt64`	`integer`	no	filter	WarrenHub canonical instrument identifier; use it for cross-dataset joins.标准术语表
4	`rank`	`UInt16`	`integer`	no	value	One-based or source-defined rank within the row's cohort.标准术语表
5	`priority_score`	`Float64`	`number`	no	value	Source-normalized field: priority score.名称推导
6	`trigger_day_count`	`UInt8`	`integer`	no	value	Count of trigger day count.名称推导
7	`detector_count`	`UInt8`	`integer`	no	value	Count of detector count.名称推导
8	`non_news_detector_count`	`UInt8`	`integer`	no	value	Count of non news detector count.名称推导
9	`trigger_sources`	`Array(String)`	`array`	no	value	Source-normalized field: trigger sources.名称推导
10	`trigger_directions`	`Array(String)`	`array`	no	value	Source-normalized field: trigger directions.名称推导
11	`first_trigger_ts`UTC_datetime	`Nullable(DateTime64(3, 'UTC'))`	`string / date-time`	yes	value	UTC timestamp for first trigger ts.名称推导
12	`latest_trigger_ts`UTC_datetime	`Nullable(DateTime64(3, 'UTC'))`	`string / date-time`	yes	value	UTC timestamp for latest trigger ts.名称推导
13	`premarket_price`	`Float64`	`number`	no	value	Price value for premarket price.名称推导
14	`premarket_vwap`	`Float64`	`number`	no	value	Source-normalized field: premarket vwap.名称推导
15	`premarket_volume`	`UInt64`	`integer`	no	value	Quantity for premarket volume.名称推导
16	`bar_start`	`String`	`string`	no	value	Source-normalized field: bar start.名称推导
17	`bar_end`	`String`	`string`	no	value	Source-normalized field: bar end.名称推导
18	`regime_label`	`LowCardinality(String)`	`string`	no	value	Source-normalized field: regime label.名称推导
19	`regime_top_strip`	`String`	`string`	no	value	Source-normalized field: regime top strip.名称推导
20	`macro_hint`	`String`	`string`	no	value	Source-normalized field: macro hint.名称推导
21	`status`	`LowCardinality(String)`	`string`	no	value	Source-normalized field: status.名称推导
22	`evidence_json`	`String`	`string`	no	value	Source-normalized field: evidence json.名称推导
23	`producer_version`	`LowCardinality(String)`	`string`	no	value	Source-normalized field: producer version.名称推导
24	`run_id`	`String`	`string`	no	value	Identifier for run.名称推导
25	`version`	`UInt64`	`integer`	no	provenance	Monotonic row version used for correction-aware reads.标准术语表
26	`is_deleted`	`UInt8`	`integer`	no	value	Soft-delete marker; public serving excludes current tombstones.标准术语表
27	`created_at`UTC_datetime	`DateTime64(3, 'UTC')`	`string / date-time`	no	value	UTC timestamp when the warehouse row was first created.标准术语表

字典口径

结构来自 Friends 只读身份可见的真实 serving relation；权限、过滤、排序、PIT 与分页来自可执行 DatasetSpec。Schema observed at 2026-10-06T18:08:52Z。