WHAT IT ANSWERS
这张表适合回答什么
- 按标的和时间检索事件或原文
- 与行情联接研究事件前后反应
DATASET · event
event.transcript_feature
这是公告、新闻与可交易事件中的“业绩电话会文本特征”数据集;一行由 instrument_id、fiscal_year、fiscal_quarter、extractor_version 定位,主要字段包括 symbol、transcript_date、published_ts、feature_available_ts、qa_cosine_sim。
published_tsWHAT IT ANSWERS
GRAIN & TIME
KEY FIELDS
instrument_idfiscal_yearfiscal_quarterextractor_versionRECOMMENDED JOINS
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。
HOW TO QUERY SAFELY
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。
MACHINE CONTRACT
instrument_id, fiscal_year, fiscal_quarter, extractor_versioninstrument_id, fiscal_year, fiscal_quarter, extractor_versioncanonical_final_activefiscal_quarter→ fiscal_quarter · integerfiscal_year→ fiscal_year · integerinstrument_id→ instrument_id · integerprovider→ provider · stringsource→ source · stringsymbol→ symbol · stringPOINT IN TIME
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
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。
curl --fail-with-body -H "Authorization: Bearer $FRIENDS_DATA_API_KEY" "https://api.alphahubs.uk/api/v1/data/datasets/event/transcript_feature/rows?limit=100&instrument_id=1"from friends_data_api import FriendsDataClient, Settings
async with FriendsDataClient(Settings.from_env()) as client:
page = await client.get_dataset_rows('event', 'transcript_feature', limit=100, filters={'instrument_id': 1})
print(page.data)
if page.has_more:
print("next cursor:", page.next_cursor)friends-data get-dataset-rows event transcript_feature --limit 100 --filter instrument_id=1{
"arguments": {
"dataset": "transcript_feature",
"filters": {
"instrument_id": 1
},
"limit": 100,
"namespace": "event"
},
"tool": "get_dataset_rows"
}FIELD DICTIONARY
descriptionSource 可在 JSON 中区分仓库注释、标准术语表和名称推导说明。
| # | 字段 | ClickHouse 类型 | JSON | 可空 | 角色 | 说明 |
|---|---|---|---|---|---|---|
| 1 | instrument_id | UInt64 | integer | no | filter, default_order, warehouse_primary_key, contract_unique_key | WarrenHub canonical instrument identifier; use it for cross-dataset joins.标准术语表 |
| 2 | symbol | String | string | no | filter | Display or trading symbol for the instrument.标准术语表 |
| 3 | fiscal_year | UInt16 | integer | no | filter, default_order, warehouse_primary_key, contract_unique_key | Source-normalized field: fiscal year.名称推导 |
| 4 | fiscal_quarter | UInt8 | integer | no | filter, default_order, warehouse_primary_key, contract_unique_key | Source-normalized field: fiscal quarter.名称推导 |
| 5 | transcript_datedate | Date | string / date | no | value | Calendar date for transcript date (YYYY-MM-DD).名称推导 |
| 6 | published_tsUTC_datetime | DateTime64(3, 'UTC') | string / date-time | no | time, provenance | UTC timestamp when the source published the record.标准术语表 |
| 7 | feature_available_tsUTC_datetime | DateTime64(3, 'UTC') | string / date-time | no | value | UTC timestamp for feature available ts.名称推导 |
| 8 | qa_cosine_sim | Nullable(Float64) | number | yes | value | Source-normalized field: qa cosine sim.名称推导 |
| 9 | n_qa_pairs | UInt32 | integer | no | value | Count of qa pairs.名称推导 |
| 10 | yoy_remarks_cosine | Nullable(Float64) | number | yes | value | Source-normalized field: yoy remarks cosine.名称推导 |
| 11 | prior_call_fiscal_year | Nullable(UInt16) | integer | yes | value | Source-normalized field: prior call fiscal year.名称推导 |
| 12 | prior_call_fiscal_quarter | Nullable(UInt8) | integer | yes | value | Source-normalized field: prior call fiscal quarter.名称推导 |
| 13 | lm_tone_net | Nullable(Float64) | number | yes | value | Source-normalized field: lm tone net.名称推导 |
| 14 | lm_tone_qa | Nullable(Float64) | number | yes | value | Source-normalized field: lm tone qa.名称推导 |
| 15 | lm_tone_prepared | Nullable(Float64) | number | yes | value | Source-normalized field: lm tone prepared.名称推导 |
| 16 | lm_pos_count | UInt32 | integer | no | value | Count of lm pos count.名称推导 |
| 17 | lm_neg_count | UInt32 | integer | no | value | Count of lm neg count.名称推导 |
| 18 | lm_uncertainty_count | UInt32 | integer | no | value | Count of lm uncertainty count.名称推导 |
| 19 | lm_litigious_count | UInt32 | integer | no | value | Count of lm litigious count.名称推导 |
| 20 | n_prepared_turns | UInt32 | integer | no | value | Count of prepared turns.名称推导 |
| 21 | n_qa_turns | UInt32 | integer | no | value | Count of qa turns.名称推导 |
| 22 | token_count | UInt32 | integer | no | value | Count of token count.名称推导 |
| 23 | extractor_version | LowCardinality(String) | string | no | default_order, warehouse_primary_key, contract_unique_key | Source-normalized field: extractor version.名称推导 |
| 24 | source | LowCardinality(String) | string | no | filter, provenance | Normalized upstream source identifier.标准术语表 |
| 25 | provider | LowCardinality(String) | string | no | filter, provenance | Upstream data provider identifier.标准术语表 |
| 26 | quality_status | LowCardinality(String) | string | no | provenance | Machine-readable data-quality classification.标准术语表 |
| 27 | visibility_tier | LowCardinality(String) | string | no | provenance | Serving visibility tier attached to the row.标准术语表 |
| 28 | ingest_run_id | String | string | no | provenance | Identifier of the ingestion run that produced the row.标准术语表 |
| 29 | version | UInt64 | integer | no | provenance | Monotonic row version used for correction-aware reads.标准术语表 |
| 30 | version_tsUTC_datetime | DateTime64(3, 'UTC') | string / date-time | no | provenance | UTC timestamp associated with the row version.标准术语表 |
| 31 | is_deleted | UInt8 | integer | no | value | Soft-delete marker; public serving excludes current tombstones.标准术语表 |
换一个字段名、类型或角色。
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。