feat: 指数成分成交汇总 + 债券/回购价格标度修复 + 美股分时 hour 修复 (0.6.5) (#6)
* feat: 指数成分成交汇总 + 债券/回购价格标度修复 + 美股分时 hour 修复 新增 - get_index_members / get_index_trade_stats / get_index_trade_stats_multi: 由成分股逐笔累加出指数的每分钟成交量与成交笔数(协议里指数本身没有笔数, trade_count 实为均价*1000) - CLI: tdx index-stats 修复 - price_decimal_digits 补齐债券/回购 4 位小数档:沪市少 100 倍、深市少 10 倍, 沪市 B 股 900xxx 改为 3 位。已用全量 51982 只标的的 SecurityInfo.decimal_point 逐条比对,零例外 - MinuteBar.hour 对美股(EX)出现 24~27:改为 hour%=24 + day_offset, 新增 datetime / trading_date 只读属性;A股/港股零行为变化 测试 217 -> 289 * chore: bump version 0.6.5 --------- Co-authored-by: openbot-coder <openbot-coder@users.noreply.github.com>
This commit is contained in:
1 parent
ca24521d2c
commit
b84a6c8a1b
15 files changed
+1129
-21
No files matched your search
+123
@@ -2,6 +2,129 @@
|
|||||||
|
|
||||||
本项目遵循 [语义化版本](https://semver.org/lang/zh-CN/)。
|
本项目遵循 [语义化版本](https://semver.org/lang/zh-CN/)。
|
||||||
|
|
||||||
|
## [0.6.5] - 2026-10-22
|
||||||
|
|
||||||
|
### 新增
|
||||||
|
|
||||||
|
- **指数成分成交汇总:一条接口拿到「某指数下全部股票的成交量与成交笔数」**
|
||||||
|
(`get_index_members` / `get_index_trade_stats` / `get_index_trade_stats_multi`)。
|
||||||
|
|
||||||
|
起因是一个绕不开的协议事实:**指数本身没有成交笔数**。指数逐笔(`sh000001`)的
|
||||||
|
`trade_count` 字段实际是「当时均价 × 1000」(2026-10-21 已验证),`get_minutes` /
|
||||||
|
`get_market_stat` / `SecurityBar` / `get_quotes` 的 160 个字段位里也都**没有**笔数
|
||||||
|
(扫描 0x122C 全字段位确认,`0x8`/`0x9` 是外盘/内盘手数)。所以「指数笔数」唯一可行
|
||||||
|
的口径是**把成分股的逐笔笔数累加**。
|
||||||
|
|
||||||
|
```python
|
||||||
|
async with TdxData() as td:
|
||||||
|
st = await td.get_index_trade_stats("000300") # 沪深300,约 10 s
|
||||||
|
for hhmm, cum in st.cum_trade_count(): # 截止该分钟的累计笔数
|
||||||
|
print(hhmm, cum)
|
||||||
|
st.minutes[0] # MinuteTradeStat(time='09:25', volume=…, trade_count=…)
|
||||||
|
st.total_volume # 手
|
||||||
|
st.total_trade_count # 笔
|
||||||
|
```
|
||||||
|
|
||||||
|
| 方法 | 作用 |
|
||||||
|
|---|---|
|
||||||
|
| `get_index_members(index, *, fields=None, count=100000)` | 指数成分股(含实时报价)。裸 6 位按**指数语义**解析:`000001`=上证指数,不是平安银行 |
|
||||||
|
| `get_index_trade_stats(index, date=None, *, concurrency=24, batch_size=50, members=None, skip_errors=True, on_progress=None)` | 成分逐笔 → 每分钟成交量/笔数;`date` 支持历史 |
|
||||||
|
| `get_index_trade_stats_multi(index, dates, *, ...)` | 多日批量(成分只请求一次),用于「近 N 日同时点均值」基准 |
|
||||||
|
|
||||||
|
**口径校验(2026-10-22 实测)**:``total_amount`` 是**成分股**合计。对「成分 = 指数
|
||||||
|
覆盖面」的指数,它与指数快照 ``amount`` 完全相等 —— 说明成分清单与量能口径一致。
|
||||||
|
|
||||||
|
| 指数 | 成分 | 成分合计 | 指数快照 | 比值 | 耗时 |
|
||||||
|
|---|---|---|---|---|---|
|
||||||
|
| `sh000016` 上证50 | 50 | 661.38 亿 | 661.34 亿 | 1.0001 | 5.8 s |
|
||||||
|
| `sh000300` 沪深300 | 300 | 2373.98 亿 | 2373.98 亿 | 1.0000 | 9.7 s |
|
||||||
|
| `sh000001` 上证指数 | 2230 | 3907.00 亿 | 4111.56 亿 | 0.9503 | 35.6 s |
|
||||||
|
|
||||||
|
历史日期同样成立(2026-10-21 收盘数据):`sh000016` 1046.29 / 1045.59 亿(1.0007)、
|
||||||
|
`sh000300` 3881.65 / 3879.30 亿(1.0006)。
|
||||||
|
|
||||||
|
单只股票级别的口径更硬:`Σ(逐笔价 × 手 × 100)` 与日线 `amount` 比值
|
||||||
|
**1.0000~1.0004**(`sh600000` / `sz000001` / `sh601398` / `bj920002` / `bj920015`
|
||||||
|
全部吻合),`Σvol` 与快照 `vol` 比值 1.0000。
|
||||||
|
|
||||||
|
CLI 同步新增:`tdx index-stats 000300 --top 10`。
|
||||||
|
|
||||||
|
**三个必须知道的限制**:
|
||||||
|
|
||||||
|
1. **深市 `399xxx` 的指数成交额是「全板块」口径**,不是样本股 —— 实测
|
||||||
|
`sz399001` 快照 5081.40 亿 = 深市 2908 只股票合计 5081.01 亿,`sz399006`
|
||||||
|
2568.60 亿 = 创业板全部股票合计。所以拿它跟 `total_amount` 比会得到 0.43 这种
|
||||||
|
无意义比值,不是数据错。**`get_index_trade_stats` 返回的始终是成分口径。**
|
||||||
|
2. **北交所(`899xxx`)没有笔数字段** —— 逐笔 `trade_count` 恒为 0,此时
|
||||||
|
`trade_count_available` 为 `False`,只有成交量可用。
|
||||||
|
3. **成分表比全市场滞后** —— 上证指数返回 2230 只,而 `get_universe("sh")` 里
|
||||||
|
股票段共 2360 只,差 130 只(新上市尚未纳入服务端板块),这正是上表
|
||||||
|
0.9503 的来源;`sh000016`/`sh000300` 这类固定样本指数不受影响。
|
||||||
|
需要精确覆盖时可自行传 `members=[...]`。
|
||||||
|
|
||||||
|
### 修复
|
||||||
|
|
||||||
|
- **债券与回购价格全线偏差:沪市少 100 倍、深市少 10 倍**(`get_quotes`)。
|
||||||
|
|
||||||
|
`price_decimal_digits` 只覆盖了「股票 2 位 / 基金 3 位」两档,债券与回购没有对应
|
||||||
|
分支,落到默认档位。它们实际是 **4 位小数**,于是被读大了两个数量级:
|
||||||
|
|
||||||
|
| 标的 | 真实价 | 修复前 | 偏差 |
|
||||||
|
|---|---|---|---|
|
||||||
|
| `sh019742` 国债 24特国01 | 110.2280 | 11022.8000 | ×100 |
|
||||||
|
| `sh110077` 可转债 洪城转债 | 199.2790 | 19927.9000 | ×100 |
|
||||||
|
| `sh204001` 回购 GC001 | 1.3850 | 138.5000 | ×100 |
|
||||||
|
| `sh900901` B股 云赛B股 | 0.5670 | 5.6700 | ×10 |
|
||||||
|
| `sz123107` 可转债 温氏转债 | 112.1890 | 1121.8900 | ×10 |
|
||||||
|
| `sz131810` 回购 R-001 | 1.3800 | 13.8000 | ×10 |
|
||||||
|
|
||||||
|
沪市 `010/018/019/020/1xx/2xx` 与深市 `10x-14x/19x/50x/52x/56x` 全部改为 4 位;
|
||||||
|
沪市 B 股 `900xxx` 另改为 3 位(与深市 `200xxx` 的 2 位不同,两市不一致)。
|
||||||
|
股票 / ETF / 指数 / 板块的行为**不变**。
|
||||||
|
|
||||||
|
**验证方式(可作为该函数的验收标准)**:拿 `pre_close`(五档快照口径,走
|
||||||
|
`price_decimal_digits`)除以上一交易日**日线收盘**(K线口径,走独立的 `/1000`
|
||||||
|
路径),比值 1.0 即两口径一致。修复后 15 类标的全数为 1.0000,零偏差。
|
||||||
|
|
||||||
|
**权威来源已确认**:证券列表里协议**自带** `SecurityInfo.decimal_point` 字段。
|
||||||
|
2026-10-22 拉取全量 **51,982 只**(沪 27,520 + 深 24,462)逐条比对,本函数与
|
||||||
|
协议字段**完全一致、零例外**——因为全量清单里没有任何一个「三位前缀」出现混合
|
||||||
|
小数位。`get_quotes` 解析时拿不到证券清单,故仍用前缀推断;需要权威值时读
|
||||||
|
`SecurityInfo.decimal_point`。
|
||||||
|
|
||||||
|
- **美股(EX 通道)分时的 `hour` 是非法时钟值**(`get_minutes("usAAPL")`)。
|
||||||
|
|
||||||
|
EX 分时的 `minutes` 字段是**自开盘起的连续分钟数**,不是「当天的分钟数」。
|
||||||
|
旧代码直接 `minutes // 60`,于是 390 条里 **240 条** 的 `hour` ≥ 24:
|
||||||
|
|
||||||
|
```
|
||||||
|
修复前: 21:30, 21:31, ..., 23:59, 24:00, 24:01, ..., 27:59 ← hour=27 不存在
|
||||||
|
修复后: 21:30, 21:31, ..., 23:59, 00:00, 00:01, ..., 03:59 ← 真实北京时间
|
||||||
|
```
|
||||||
|
|
||||||
|
任何 `datetime(...)` 拼接都会因 `hour=27` 抛 `ValueError`,按小时聚合也会把
|
||||||
|
凌晨段错误地归到同一个自然日。
|
||||||
|
|
||||||
|
**修法**:`MinuteBar.__post_init__` 把越界的 `hour` 拆成
|
||||||
|
「`hour %= 24` + `day_offset += hour // 24`」,并新增两个只读属性:
|
||||||
|
|
||||||
|
| 属性 | 用途 |
|
||||||
|
|---|---|
|
||||||
|
| `MinuteBar.datetime` | 该分时点的时刻(北京时间,已含跨日偏移) |
|
||||||
|
| `MinuteBar.trading_date` | 归属的交易日 `YYYY-MM-DD`。美股一个交易日横跨两个北京自然日(21:30~次日 04:00),**按交易日聚合必须用它**,与 `SecurityBar.trading_date` 口径一致 |
|
||||||
|
|
||||||
|
A股(合成时间 09:30~15:00)/ 港股(09:30~16:00)/ 期货的 `hour` 本就合法
|
||||||
|
→ **零行为变化**,`day_offset` 恒为 0。
|
||||||
|
|
||||||
|
实测(2026-10-22):`usAAPL` / `usTSLA` / `usNVDA` 各 390 条,修复后
|
||||||
|
`hour` 全在 0~23、时间轴严格单调、`trading_date` 收敛为单一交易日;
|
||||||
|
`hk00700` 90 条无任何变化。
|
||||||
|
|
||||||
|
### 说明
|
||||||
|
|
||||||
|
- 新增回归测试:债券/回购/B股小数位 ×29、美股分时跨日 ×6、指数成分成交汇总 ×34、
|
||||||
|
全接口 ×1,测试总数 **217 → 289**。
|
||||||
|
|
||||||
## [0.6.4] - 2026-10-22
|
## [0.6.4] - 2026-10-22
|
||||||
|
|
||||||
### 修复
|
### 修复
|
||||||
|
|||||||
@@ -149,6 +149,8 @@ await td.get_bars("cffex:IFL0", KlinePeriod.DAY, count=700) # 期货合约
|
|||||||
| 除权除息 / 财务快照 | `get_xdxr(symbols)` / `get_finance(symbols)` |
|
| 除权除息 / 财务快照 | `get_xdxr(symbols)` / `get_finance(symbols)` |
|
||||||
| F10 公司资料 | `get_f10(symbol, section)` |
|
| F10 公司资料 | `get_f10(symbol, section)` |
|
||||||
| 板块列表 / 成分股 / 涨跌排行 | `get_boards` / `get_board_members` / `get_board_ranking` |
|
| 板块列表 / 成分股 / 涨跌排行 | `get_boards` / `get_board_members` / `get_board_ranking` |
|
||||||
|
| **指数的成交量 + 成交笔数**(每分钟曲线) | `get_index_trade_stats(index, date)` → `.cum_trade_count()` |
|
||||||
|
| **指数成分股** | `get_index_members(index)`(裸 `000001` 按指数语义解析) |
|
||||||
| 个股所属板块 | `get_board_of(symbols)` |
|
| 个股所属板块 | `get_board_of(symbols)` |
|
||||||
| 集合竞价 / 异动 / 资金流 | `get_auction` / `get_unusual` / `get_capital_flow` |
|
| 集合竞价 / 异动 / 资金流 | `get_auction` / `get_unusual` / `get_capital_flow` |
|
||||||
| 个股特征快照 | `get_snapshot(symbols)` |
|
| 个股特征快照 | `get_snapshot(symbols)` |
|
||||||
@@ -167,7 +169,8 @@ await td.get_bars("cffex:IFL0", KlinePeriod.DAY, count=700) # 期货合约
|
|||||||
- **K 线时间戳统一为北京时间**(v0.6.2 起)。美股原始报文里日期是美东交易日、时分是北京时间(两个时区拼在一起),库已自动归一化,`bar.datetime` 可直接排序 / 比较 / 过滤。但**美股一个交易日横跨两个北京自然日**,按交易日聚合(分钟线合成日线、日内统计)必须用 `bar.trading_date`。
|
- **K 线时间戳统一为北京时间**(v0.6.2 起)。美股原始报文里日期是美东交易日、时分是北京时间(两个时区拼在一起),库已自动归一化,`bar.datetime` 可直接排序 / 比较 / 过滤。但**美股一个交易日横跨两个北京自然日**,按交易日聚合(分钟线合成日线、日内统计)必须用 `bar.trading_date`。
|
||||||
- **`count` 很大时会自动多连接并发分页**,无需手动翻页。但标准通道 K 线 `count < 100` 部分服务器会返回空——内部已强制 `min_page=100` 规避。
|
- **`count` 很大时会自动多连接并发分页**,无需手动翻页。但标准通道 K 线 `count < 100` 部分服务器会返回空——内部已强制 `min_page=100` 规避。
|
||||||
- **`get_quotes` 一次 >80 只自动切批**;走 MAC 字段位通道(`fields=`)时上限为 80 只/次。
|
- **`get_quotes` 一次 >80 只自动切批**;走 MAC 字段位通道(`fields=`)时上限为 80 只/次。
|
||||||
- **历史逐笔没有笔数字段**,只有当日逐笔的 `trade_count` 有效。
|
- **指数本身没有成交笔数**。`get_ticks("sh000001")` 的 `trade_count` 是「当时均价×1000」、`vol` 单位是百元(成交额/100)。要笔数只能用 `get_index_trade_stats(index)`(成分股逐笔累加,2230 只约 36-60 s)。北交所成分无笔数字段,该方法会返回 `trade_count_available=False`。
|
||||||
|
- **历史逐笔有笔数**(MAC `0x122F` 支持历史日期,实测可回溯最近 5 个交易日;更早返回空)。
|
||||||
- 板块文件(`block_*.dat`)/ 行业(`tdxhy.cfg`)/ 财报(`gpcw*.zip`)是**服务器上的静态文件**,解析后是普通 dataclass,**落 JSON 后完全脱网可用**——适合做离线快照。
|
- 板块文件(`block_*.dat`)/ 行业(`tdxhy.cfg`)/ 财报(`gpcw*.zip`)是**服务器上的静态文件**,解析后是普通 dataclass,**落 JSON 后完全脱网可用**——适合做离线快照。
|
||||||
|
|
||||||
## 陷阱清单
|
## 陷阱清单
|
||||||
@@ -185,6 +188,8 @@ await td.get_bars("cffex:IFL0", KlinePeriod.DAY, count=700) # 期货合约
|
|||||||
| 循环里逐个 `await get_bars` | 慢 | 用列表传多个标的,自动并发 |
|
| 循环里逐个 `await get_bars` | 慢 | 用列表传多个标的,自动并发 |
|
||||||
| 港股/美股五档 bid/ask | 没有(协议不支持) | `fields` 字段位通道有 name 但无五档 |
|
| 港股/美股五档 bid/ask | 没有(协议不支持) | `fields` 字段位通道有 name 但无五档 |
|
||||||
| 美股分钟线用 `datetime.date()` 分组做日聚合 | 一个交易日被拆成两天(美股跨北京零点) | 用 `bar.trading_date`;`bar.datetime` 是归一化后的北京时间 |
|
| 美股分钟线用 `datetime.date()` 分组做日聚合 | 一个交易日被拆成两天(美股跨北京零点) | 用 `bar.trading_date`;`bar.datetime` 是归一化后的北京时间 |
|
||||||
|
| 拿指数逐笔的 `trade_count` 当笔数 | 拿到的是「均价×1000」,数值荒谬 | 用 `get_index_trade_stats(index)` 由成分累加 |
|
||||||
|
| 拿深市 `399xxx` 的 `amount` 校验成分合计 | 比值 0.43,误以为数据错 | 深市指数 amount 是**全板块**口径(`sz399001`=深市全部 5081 亿),不是样本股 |
|
||||||
| 依赖实时性做信号 | 数据是 15 分钟前的 | 换券商实时源 |
|
| 依赖实时性做信号 | 数据是 15 分钟前的 | 换券商实时源 |
|
||||||
| 手动改包内 `servers.json` / 写死 IP | 容器只读、升级被覆盖、快节点捞不回来 | 日常靠内置的**当日测速缓存**自动选路;要强制指定就用 `PYPYTDX_DATA_{STANDARD,MAC,EX}_HOSTS` 环境变量 |
|
| 手动改包内 `servers.json` / 写死 IP | 容器只读、升级被覆盖、快节点捞不回来 | 日常靠内置的**当日测速缓存**自动选路;要强制指定就用 `PYPYTDX_DATA_{STANDARD,MAC,EX}_HOSTS` 环境变量 |
|
||||||
|
|
||||||
|
|||||||
@@ -81,6 +81,9 @@ from pytdxdata import (
|
|||||||
| `get_boards(kind="all", *, count=10000)` | `list[BoardInfo]` | 板块列表 |
|
| `get_boards(kind="all", *, count=10000)` | `list[BoardInfo]` | 板块列表 |
|
||||||
| `get_board_members(board, *, fields=None, count=100000)` | `list[SecurityQuote]` | 板块成分股报价(归一模型) |
|
| `get_board_members(board, *, fields=None, count=100000)` | `list[SecurityQuote]` | 板块成分股报价(归一模型) |
|
||||||
| `get_board_of(symbols)` | `list[BelongBoard]` | 个股所属板块 |
|
| `get_board_of(symbols)` | `list[BelongBoard]` | 个股所属板块 |
|
||||||
|
| `get_index_members(index, *, fields=None, count=100000)` | `list[SecurityQuote]` | **指数成分股**报价。裸 6 位按**指数语义**解析(`000001`=上证指数,不是平安银行):`000/999/880`→沪、`399`→深、`899`→北交所 |
|
||||||
|
| `get_index_trade_stats(index, date=None, *, concurrency=24, batch_size=50, members=None, skip_errors=True, on_progress=None)` | `IndexTradeStats` | **指数成分逐笔汇总**成每分钟成交量/成交笔数。协议里指数本身没有笔数(`trade_count` 是均价×1000),故由成分股逐笔累加;`.cumulative()` 取「截止该分钟累计」,`.cum_trade_count()` 直接给 `[(HH:MM, 累计笔数)]`。2230 只(上证)约 36-60 s |
|
||||||
|
| `get_index_trade_stats_multi(index, dates, *, ...)` | `list[IndexTradeStats]` | 多日批量版(成分只请求一次),用于「近 N 日同时点累计均值」基准 |
|
||||||
|
|
||||||
### 2.2 K线(旧接口,共 6 个) {#legacy-kline}
|
### 2.2 K线(旧接口,共 6 个) {#legacy-kline}
|
||||||
|
|
||||||
@@ -199,7 +202,7 @@ from pytdxdata import (
|
|||||||
| `SecurityQuote` | price/pre_close/OHLC/vol/amount/s_vol/b_vol/`bid[5]`/`ask[5]`/server_time/trading_status/cur_vol/rise_speed/limit_up/limit_down/decimal_point/open_amount |
|
| `SecurityQuote` | price/pre_close/OHLC/vol/amount/s_vol/b_vol/`bid[5]`/`ask[5]`/server_time/trading_status/cur_vol/rise_speed/limit_up/limit_down/decimal_point/open_amount |
|
||||||
| `SecurityInfo` | market/code/name/volunit/decimal_point/pre_close |
|
| `SecurityInfo` | market/code/name/volunit/decimal_point/pre_close |
|
||||||
| `TransactionRecord` | market/code/time/price/vol/`trade_count`/bs_flag |
|
| `TransactionRecord` | market/code/time/price/vol/`trade_count`/bs_flag |
|
||||||
| `MinuteBar` | price/vol/avg_price/hour/minute |
|
| `MinuteBar` | price/vol/avg_price/hour/minute/day_offset + `datetime`/`trading_date` 只读属性(hour 恒为合法时钟值 0~23;美股一个交易日跨两个北京自然日,按交易日聚合用 `trading_date`) |
|
||||||
| `XdxrRecord` | 分红送转/配股/缩股/行权 + 股本变动 + `date` 属性 |
|
| `XdxrRecord` | 分红送转/配股/缩股/行权 + 股本变动 + `date` 属性 |
|
||||||
| `FinanceRecord` | 股本/资产/负债/营收/利润等 36 个字段(万元/万股) |
|
| `FinanceRecord` | 股本/资产/负债/营收/利润等 36 个字段(万元/万股) |
|
||||||
| `BoardInfo` | 板块 + 领涨股信息 |
|
| `BoardInfo` | 板块 + 领涨股信息 |
|
||||||
|
|||||||
+1
-1
@@ -48,7 +48,7 @@ asyncio.run(main())
|
|||||||
- **K 线**:`get_bars`(自动并发分页、支持复权与日期区间、多标的并发)
|
- **K 线**:`get_bars`(自动并发分页、支持复权与日期区间、多标的并发)
|
||||||
- **技术指标**:`get_indicators`(K线 + MACD/KDJ/RSI/BOLL/MA/EMA)
|
- **技术指标**:`get_indicators`(K线 + MACD/KDJ/RSI/BOLL/MA/EMA)
|
||||||
- **报价**:`get_quotes`(五档,>80 只自动切批)、`get_snapshot`(个股特征)、`get_market_stat`
|
- **报价**:`get_quotes`(五档,>80 只自动切批)、`get_snapshot`(个股特征)、`get_market_stat`
|
||||||
- **逐笔 / 分时**:`get_ticks`(含 `trade_count`,支持历史日期,返回按时间升序;`vol`/`trade_count` 为约 3 秒切片增量而非累计;已剔除 `bs_flag=8` 竞价委托;**指数**的 `vol` 是百元、`trade_count` 是均价×1000)、`get_minutes`(`sampling=True` 取缩略采样)
|
- **逐笔 / 分时**:`get_ticks`(含 `trade_count`,支持历史日期,返回按时间升序;`vol`/`trade_count` 为约 3 秒切片增量而非累计;已剔除 `bs_flag=8` 竞价委托;**指数**的 `vol` 是百元、`trade_count` 是均价×1000)、`get_minutes`(`sampling=True` 取缩略采样;`hour` 恒为合法时钟值 0~23,美股一个交易日跨两个北京自然日,按交易日聚合用 `MinuteBar.trading_date`、取时刻用 `MinuteBar.datetime`)
|
||||||
- **财务 / 除权**:`get_finance`(最新快照)、`get_xdxr`、`get_price_limits`(本地计算涨跌停)
|
- **财务 / 除权**:`get_finance`(最新快照)、`get_xdxr`、`get_price_limits`(本地计算涨跌停)
|
||||||
- **F10**:`get_f10`(公司资料目录 + 正文)
|
- **F10**:`get_f10`(公司资料目录 + 正文)
|
||||||
- **板块**:`get_boards`、`get_board_members`、`get_board_of`、`get_board_summary`、
|
- **板块**:`get_boards`、`get_board_members`、`get_board_of`、`get_board_summary`、
|
||||||
|
|||||||
+1
-1
@@ -1,6 +1,6 @@
|
|||||||
[project]
|
[project]
|
||||||
name = "pytdxdata"
|
name = "pytdxdata"
|
||||||
version = "0.6.4"
|
version = "0.6.5"
|
||||||
description = "通达信(TongDaXin)行情数据客户端:动态连接池 + 双层TTL缓存 + 3连接并发分页,纯asyncio零运行时依赖,支持A股/港股/美股/期货/期权"
|
description = "通达信(TongDaXin)行情数据客户端:动态连接池 + 双层TTL缓存 + 3连接并发分页,纯asyncio零运行时依赖,支持A股/港股/美股/期货/期权"
|
||||||
readme = "README.md"
|
readme = "README.md"
|
||||||
requires-python = ">=3.11"
|
requires-python = ">=3.11"
|
||||||
|
|||||||
@@ -0,0 +1,108 @@
|
|||||||
|
"""可复现核查:指数成分逐笔汇总的口径是否与官方日线一致。
|
||||||
|
|
||||||
|
用法::
|
||||||
|
|
||||||
|
python scripts/verify_index_stats.py # 当日
|
||||||
|
python scripts/verify_index_stats.py --date 20261021 # 历史某日
|
||||||
|
python scripts/verify_index_stats.py --indices 000001 000300
|
||||||
|
|
||||||
|
两段校验:
|
||||||
|
|
||||||
|
1. **单股级**:``Σ(逐笔价 × 手 × 100)`` 对**当日日线 amount**。比值 1.0000~1.0002
|
||||||
|
说明逐笔没漏条、价格与单位都对(``bj`` 也行)。
|
||||||
|
2. **指数级**:成分股合计成交额对**该指数的日线 amount**。相等说明成分清单与
|
||||||
|
指数的量能口径一致;明显偏小则是成分表滞后(新上市未纳入)。
|
||||||
|
|
||||||
|
⚠️ **深市 `399xxx` 的指数 amount 是「全板块」口径**,不是样本股:实测
|
||||||
|
`sz399001` 5081.40 亿 = 深市 2908 只股票合计,`sz399006` 2568.60 亿 = 创业板
|
||||||
|
全部股票合计。所以这两个指数的比值天然远离 1(约 0.43),**不是数据错**。
|
||||||
|
|
||||||
|
为什么不用快照做基准:快照只有「此刻」,历史日期对不上。日线 amount 两者都能给。
|
||||||
|
|
||||||
|
注意 ``bj``(北交所)逐笔**没有笔数字段**(``trade_count`` 恒 0),脚本会打出
|
||||||
|
``有笔数`` 列便于识别。
|
||||||
|
"""
|
||||||
|
from __future__ import annotations
|
||||||
|
|
||||||
|
import argparse
|
||||||
|
import asyncio
|
||||||
|
from datetime import date as _date
|
||||||
|
|
||||||
|
|
||||||
|
def _ymd(bar) -> int:
|
||||||
|
return bar.year * 10000 + bar.month * 100 + bar.day
|
||||||
|
|
||||||
|
|
||||||
|
async def _daily_amount(td, symbol: str, ymd: int) -> float | None:
|
||||||
|
from pytdxdata import KlinePeriod
|
||||||
|
|
||||||
|
bars = await td.get_bars(symbol, KlinePeriod.DAY, count=10)
|
||||||
|
for b in bars:
|
||||||
|
if _ymd(b) == ymd:
|
||||||
|
return b.amount
|
||||||
|
return None
|
||||||
|
|
||||||
|
|
||||||
|
async def check_symbols(td, symbols: list[str], ymd: int) -> None:
|
||||||
|
print(f"## 单股级({ymd})\n")
|
||||||
|
print("| 标的 | Σ(p×手×100) / 日线amount | 笔数 |")
|
||||||
|
print("|---|---|---|")
|
||||||
|
for sym in symbols:
|
||||||
|
try:
|
||||||
|
ticks = await td.get_ticks(sym, date=ymd)
|
||||||
|
ref = await _daily_amount(td, sym, ymd)
|
||||||
|
except Exception as exc: # noqa: BLE001
|
||||||
|
print(f"| {sym} | ERR {type(exc).__name__} | - |")
|
||||||
|
continue
|
||||||
|
if not ticks:
|
||||||
|
print(f"| {sym} | 无逐笔(非交易日?) | - |")
|
||||||
|
continue
|
||||||
|
sa = sum(t.price * t.vol * 100 for t in ticks)
|
||||||
|
tc = sum(t.trade_count for t in ticks)
|
||||||
|
ratio = sa / ref if ref else 0.0
|
||||||
|
print(f"| {sym} | {ratio:.4f} | {tc:,} |")
|
||||||
|
print()
|
||||||
|
|
||||||
|
|
||||||
|
async def check_indices(td, indices: list[str], ymd: int, day: int | None) -> None:
|
||||||
|
print(f"## 指数级({ymd})\n")
|
||||||
|
print("| 指数 | 成分 | 有成交 | 有笔数 | 成分合计 | 日线amount | 比值 | 耗时 |")
|
||||||
|
print("|---|---|---|---|---|---|---|---|")
|
||||||
|
for idx in indices:
|
||||||
|
try:
|
||||||
|
st = await td.get_index_trade_stats(idx, day)
|
||||||
|
ref = await _daily_amount(td, st.index, ymd)
|
||||||
|
except Exception as exc: # noqa: BLE001
|
||||||
|
print(f"| {idx} | ERR {type(exc).__name__}: {exc} | - | - | - | - | - | - |")
|
||||||
|
continue
|
||||||
|
ratio = st.total_amount / ref if ref else 0.0
|
||||||
|
print(f"| {st.index} | {st.member_count} | {st.active_members} | "
|
||||||
|
f"{st.trade_count_members} | {st.total_amount / 1e8:,.2f} 亿 | "
|
||||||
|
f"{(ref or 0) / 1e8:,.2f} 亿 | {ratio:.4f} | {st.elapsed:.1f}s |")
|
||||||
|
print()
|
||||||
|
print("提示:`399xxx` 的比值天然远离 1 —— 深市指数的 amount 字段是全板块口径。")
|
||||||
|
print()
|
||||||
|
|
||||||
|
|
||||||
|
async def main() -> None:
|
||||||
|
ap = argparse.ArgumentParser(description=__doc__)
|
||||||
|
ap.add_argument("--symbols", nargs="*",
|
||||||
|
default=["sh600000", "sz000001", "sh601398"])
|
||||||
|
ap.add_argument("--indices", nargs="*", default=["000016", "000300"])
|
||||||
|
ap.add_argument("--date", default=None, help="YYYYMMDD(缺省=当日)")
|
||||||
|
ap.add_argument("--skip-symbols", action="store_true")
|
||||||
|
args = ap.parse_args()
|
||||||
|
|
||||||
|
from pytdxdata import TdxData
|
||||||
|
|
||||||
|
ymd = int(args.date) if args.date else int(_date.today().strftime("%Y%m%d"))
|
||||||
|
day = int(args.date) if args.date else None
|
||||||
|
async with TdxData(cache_dir=None, timeout=10.0) as td:
|
||||||
|
if not args.skip_symbols:
|
||||||
|
await check_symbols(td, args.symbols, ymd)
|
||||||
|
if args.indices:
|
||||||
|
await check_indices(td, args.indices, ymd, day)
|
||||||
|
|
||||||
|
|
||||||
|
if __name__ == "__main__":
|
||||||
|
asyncio.run(main())
|
||||||
@@ -7,16 +7,18 @@ from .models import (
|
|||||||
Adjust,
|
Adjust,
|
||||||
ExMarket,
|
ExMarket,
|
||||||
FinanceRecord,
|
FinanceRecord,
|
||||||
|
IndexTradeStats,
|
||||||
KlinePeriod,
|
KlinePeriod,
|
||||||
Market,
|
Market,
|
||||||
MinuteBar,
|
MinuteBar,
|
||||||
|
MinuteTradeStat,
|
||||||
SecurityBar,
|
SecurityBar,
|
||||||
SecurityQuote,
|
SecurityQuote,
|
||||||
TransactionRecord,
|
TransactionRecord,
|
||||||
XdxrRecord,
|
XdxrRecord,
|
||||||
)
|
)
|
||||||
|
|
||||||
__version__ = "0.6.4"
|
__version__ = "0.6.5"
|
||||||
__all__ = [
|
__all__ = [
|
||||||
"TdxData",
|
"TdxData",
|
||||||
"Market",
|
"Market",
|
||||||
@@ -27,6 +29,8 @@ __all__ = [
|
|||||||
"SecurityQuote",
|
"SecurityQuote",
|
||||||
"TransactionRecord",
|
"TransactionRecord",
|
||||||
"MinuteBar",
|
"MinuteBar",
|
||||||
|
"MinuteTradeStat",
|
||||||
|
"IndexTradeStats",
|
||||||
"XdxrRecord",
|
"XdxrRecord",
|
||||||
"FinanceRecord",
|
"FinanceRecord",
|
||||||
"__version__",
|
"__version__",
|
||||||
|
|||||||
@@ -23,7 +23,9 @@ from __future__ import annotations
|
|||||||
import asyncio
|
import asyncio
|
||||||
import logging
|
import logging
|
||||||
import re
|
import re
|
||||||
|
import time
|
||||||
import warnings
|
import warnings
|
||||||
|
from collections.abc import Callable
|
||||||
from datetime import date as _date, datetime
|
from datetime import date as _date, datetime
|
||||||
from pathlib import Path
|
from pathlib import Path
|
||||||
from typing import Sequence
|
from typing import Sequence
|
||||||
@@ -45,11 +47,13 @@ from .models import (
|
|||||||
FinancialFileInfo,
|
FinancialFileInfo,
|
||||||
FinancialRecord,
|
FinancialRecord,
|
||||||
IndicatorSet,
|
IndicatorSet,
|
||||||
|
IndexTradeStats,
|
||||||
IndustryInfo,
|
IndustryInfo,
|
||||||
KlinePeriod,
|
KlinePeriod,
|
||||||
Market,
|
Market,
|
||||||
MemberQuote,
|
MemberQuote,
|
||||||
MinuteBar,
|
MinuteBar,
|
||||||
|
MinuteTradeStat,
|
||||||
SecurityBar,
|
SecurityBar,
|
||||||
SecurityInfo,
|
SecurityInfo,
|
||||||
SecurityQuote,
|
SecurityQuote,
|
||||||
@@ -1291,6 +1295,238 @@ class TdxData:
|
|||||||
result.sort(key=lambda x: x["change_pct"], reverse=True)
|
result.sort(key=lambda x: x["change_pct"], reverse=True)
|
||||||
return result[:top_n]
|
return result[:top_n]
|
||||||
|
|
||||||
|
# ================================================================== #
|
||||||
|
# 七之二、指数成分成交汇总(成分股逐笔 → 每分钟成交量 / 笔数)
|
||||||
|
# ================================================================== #
|
||||||
|
|
||||||
|
@staticmethod
|
||||||
|
def _index_pair(index: str) -> tuple[int, str]:
|
||||||
|
"""解析指数标的 → ``(市场号, 6 位代码)``。
|
||||||
|
|
||||||
|
与 :func:`~pytdxdata.symbols.parse_symbol` 的关键区别:**裸 6 位数字
|
||||||
|
一律按指数语义解析**,不做「``000xxx`` → 深市股票」的推断 ——
|
||||||
|
在指数接口里 ``000001`` 是**上证指数**,不是平安银行 ``sz000001``。
|
||||||
|
|
||||||
|
判定规则(与通达信板块码一致)::
|
||||||
|
|
||||||
|
000xxx / 999xxx / 880xxx -> 沪市(上证指数 / A股指数 / 板块指数)
|
||||||
|
399xxx -> 深市(深证成指 / 创业板指)
|
||||||
|
899xxx -> 北交所(北证50)
|
||||||
|
|
||||||
|
显式前缀 ``sh``/``sz``/``bj`` 优先,例如 ``sz399001``、``sh000300``。
|
||||||
|
"""
|
||||||
|
s = str(index).strip().lower().replace(".", "").replace(":", "")
|
||||||
|
if not s:
|
||||||
|
raise ValueError("指数标的不能为空")
|
||||||
|
prefix_table = {"sh": 1, "sz": 0, "bj": 2}
|
||||||
|
if s[:2] in prefix_table:
|
||||||
|
code = s[2:]
|
||||||
|
if not (code.isdigit() and len(code) == 6):
|
||||||
|
raise ValueError(f"非法指数标的: {index!r}(前缀后需 6 位数字)")
|
||||||
|
return prefix_table[s[:2]], code
|
||||||
|
if not (s.isdigit() and len(s) == 6):
|
||||||
|
raise ValueError(f"非法指数标的: {index!r}(需 6 位代码或 sh/sz/bj 前缀)")
|
||||||
|
if s.startswith(("000", "999", "880")):
|
||||||
|
return 1, s
|
||||||
|
if s.startswith("399"):
|
||||||
|
return 0, s
|
||||||
|
if s.startswith("899"):
|
||||||
|
return 2, s
|
||||||
|
raise ValueError(f"无法判定指数市场: {index!r}(不是 A股指数代码段)")
|
||||||
|
|
||||||
|
async def get_index_members(self, index: str, *, count: int = 100000,
|
||||||
|
fields: Sequence[int] | None = None
|
||||||
|
) -> list[SecurityQuote]:
|
||||||
|
"""指数**成分股**(含实时报价)。
|
||||||
|
|
||||||
|
走 MAC ``0x122C`` 板块成分命令,指数代码映射为服务端板块码:``000300`` →
|
||||||
|
31300、``399001`` → 30001、``899050`` → 32050(见 :meth:`_convert_board_code`)。
|
||||||
|
实测(2026-10-22):``sh000001`` 2230 只(沪市全部)、``sh000300`` 300 只、
|
||||||
|
``sh000016`` 50 只、``sh000905`` 500 只、``sh000852`` 1000 只、
|
||||||
|
``sz399006`` 100 只、``bj899050`` 50 只。
|
||||||
|
|
||||||
|
参数:
|
||||||
|
index: 指数代码,可裸写或带前缀(``"000300"`` / ``"sh000300"``)。
|
||||||
|
count: 最多取的成分数(分页上限,默认足够覆盖全部)。
|
||||||
|
fields: 附加报价字段位(同 :meth:`get_board_members`)。
|
||||||
|
|
||||||
|
注意:返回的是**当前**成分,指数调仓后历史日期不适用。
|
||||||
|
|
||||||
|
用法::
|
||||||
|
|
||||||
|
members = await td.get_index_members("000300")
|
||||||
|
syms = [m.symbol for m in members] # ['sz000001', ...]
|
||||||
|
"""
|
||||||
|
_market, code = self._index_pair(index)
|
||||||
|
board_code = self._convert_board_code(code)
|
||||||
|
bits = [int(b) for b in fields] if fields else list(_QUOTE_BITS)
|
||||||
|
items: list[SecurityQuote] = []
|
||||||
|
start = 0
|
||||||
|
while start < count:
|
||||||
|
batch = await self._execute_mac(MacBoardMembersQuotesCmd(
|
||||||
|
board_code, sort_type=0, start=start,
|
||||||
|
page_size=_QUOTE_PAGE, bits=bits))
|
||||||
|
if not batch:
|
||||||
|
break
|
||||||
|
items.extend(_quote_from_member(m) for m in batch)
|
||||||
|
start += len(batch)
|
||||||
|
if len(batch) < _QUOTE_PAGE:
|
||||||
|
break
|
||||||
|
return items[:count]
|
||||||
|
|
||||||
|
async def get_index_trade_stats(
|
||||||
|
self,
|
||||||
|
index: str,
|
||||||
|
date: int | None = None,
|
||||||
|
*,
|
||||||
|
concurrency: int = 24,
|
||||||
|
batch_size: int = 50,
|
||||||
|
members: Sequence[str] | None = None,
|
||||||
|
skip_errors: bool = True,
|
||||||
|
on_progress: Callable[[int, int], None] | None = None,
|
||||||
|
) -> IndexTradeStats:
|
||||||
|
"""按指数**全部成分股**的逐笔汇总出每分钟成交量与成交笔数。
|
||||||
|
|
||||||
|
为什么必须绕这一圈:**通达信协议里指数本身没有成交笔数** —— 指数逐笔的
|
||||||
|
``trade_count`` 是「当时均价 × 1000」,不是笔数(见
|
||||||
|
:class:`TransactionRecord` 类注释)。所以「指数笔数」只能用成分股逐笔累加。
|
||||||
|
|
||||||
|
口径校验:``total_amount`` 是**成分股**合计。对「成分 = 指数覆盖面」的指数
|
||||||
|
它与指数快照 ``amount`` 完全相等(``sh000016`` 661.34 亿、``sh000300``
|
||||||
|
2373.98 亿,比值 1.0000);但深市 ``399xxx`` 的指数 ``amount`` 是**全板块**
|
||||||
|
口径(``sz399001`` = 深市全部股票 5081 亿、``sz399006`` = 创业板全部),
|
||||||
|
两者不可直接比。详见 :class:`IndexTradeStats`。
|
||||||
|
|
||||||
|
参数:
|
||||||
|
index: 指数代码(``"000300"`` / ``"sh000300"``,规则见 :meth:`_index_pair`)。
|
||||||
|
date: ``None`` = 当日;``YYYYMMDD`` = 历史某日(通达信逐笔仅保留最近
|
||||||
|
若干交易日,实测 2026-10-22 可回溯到 10-14,更早返回空)。
|
||||||
|
concurrency: 逐笔并发数(受 MAC 连接池上限约束,池满自动排队)。
|
||||||
|
batch_size: 每批处理的成分数 —— 批内并发拉取、批间串行聚合,
|
||||||
|
用于压住内存峰值(全量 tick 不驻留)。
|
||||||
|
members: 直接给成分股列表(跳过 :meth:`get_index_members`,
|
||||||
|
多日循环时可复用一次成分请求)。
|
||||||
|
skip_errors: 单只成分失败是否忽略(默认忽略并计入 ``failed_members``)。
|
||||||
|
on_progress: 进度回调 ``(已处理, 总数)``,每批结束调用一次。
|
||||||
|
|
||||||
|
返回:
|
||||||
|
:class:`IndexTradeStats`;``minutes`` 是**每分钟增量**,
|
||||||
|
「当日截止该分钟累计」用 ``.cumulative()`` / ``.cum_trade_count()``。
|
||||||
|
|
||||||
|
性能(实测 2026-10-22,盘中):
|
||||||
|
|
||||||
|
====================== ========== ============
|
||||||
|
指数 成分数 耗时
|
||||||
|
====================== ========== ============
|
||||||
|
``sh000016`` 上证50 50 ~5 s
|
||||||
|
``sh000300`` 沪深300 300 ~10 s
|
||||||
|
``sh000001`` 上证指数 2230 ~60-90 s
|
||||||
|
====================== ========== ============
|
||||||
|
|
||||||
|
**北交所成分没有笔数**(实测 ``trade_count`` 恒 0),此时
|
||||||
|
``trade_count_available`` 为 ``False``,只有成交量可用。
|
||||||
|
|
||||||
|
用法::
|
||||||
|
|
||||||
|
st = await td.get_index_trade_stats("000300")
|
||||||
|
for hhmm, cum in st.cum_trade_count():
|
||||||
|
print(hhmm, cum)
|
||||||
|
"""
|
||||||
|
started = time.perf_counter()
|
||||||
|
market, code = self._index_pair(index)
|
||||||
|
idx_text = format_symbol(market, code)
|
||||||
|
|
||||||
|
if members is None:
|
||||||
|
members = [m.symbol for m in await self.get_index_members(code)]
|
||||||
|
syms = [s for s in members if s]
|
||||||
|
total = len(syms)
|
||||||
|
if total == 0:
|
||||||
|
raise ValueError(f"指数 {index!r} 没有取到成分股")
|
||||||
|
|
||||||
|
vol_by_min: dict[str, float] = {}
|
||||||
|
tc_by_min: dict[str, int] = {}
|
||||||
|
amt_by_min: dict[str, float] = {}
|
||||||
|
active: set[str] = set()
|
||||||
|
with_tc: set[str] = set()
|
||||||
|
failed = 0
|
||||||
|
step = max(1, int(batch_size))
|
||||||
|
done = 0
|
||||||
|
for i in range(0, total, step):
|
||||||
|
batch = syms[i:i + step]
|
||||||
|
try:
|
||||||
|
ticks = await self.get_ticks(batch, date=date, concurrency=concurrency)
|
||||||
|
except Exception:
|
||||||
|
if not skip_errors:
|
||||||
|
raise
|
||||||
|
# 整批失败 → 退化为逐只重试,避免一只坏股拖垮一批
|
||||||
|
ticks = []
|
||||||
|
for one in batch:
|
||||||
|
try:
|
||||||
|
ticks.extend(await self.get_ticks(one, date=date))
|
||||||
|
except Exception:
|
||||||
|
failed += 1
|
||||||
|
for t in ticks:
|
||||||
|
hhmm = t.time[:5]
|
||||||
|
vol_by_min[hhmm] = vol_by_min.get(hhmm, 0.0) + t.vol
|
||||||
|
tc_by_min[hhmm] = tc_by_min.get(hhmm, 0) + t.trade_count
|
||||||
|
amt_by_min[hhmm] = amt_by_min.get(hhmm, 0.0) + t.price * t.vol * 100.0
|
||||||
|
active.add(t.symbol)
|
||||||
|
if t.trade_count:
|
||||||
|
with_tc.add(t.symbol)
|
||||||
|
done += len(batch)
|
||||||
|
if on_progress is not None:
|
||||||
|
on_progress(min(done, total), total)
|
||||||
|
|
||||||
|
minutes = [
|
||||||
|
MinuteTradeStat(time=k, volume=vol_by_min[k],
|
||||||
|
trade_count=tc_by_min[k], amount=amt_by_min[k])
|
||||||
|
for k in sorted(vol_by_min)
|
||||||
|
]
|
||||||
|
return IndexTradeStats(
|
||||||
|
index=idx_text,
|
||||||
|
date=date,
|
||||||
|
member_count=total,
|
||||||
|
active_members=len(active),
|
||||||
|
trade_count_members=len(with_tc),
|
||||||
|
minutes=minutes,
|
||||||
|
total_volume=sum(vol_by_min.values()),
|
||||||
|
total_trade_count=sum(tc_by_min.values()),
|
||||||
|
total_amount=sum(amt_by_min.values()),
|
||||||
|
failed_members=failed,
|
||||||
|
elapsed=time.perf_counter() - started,
|
||||||
|
)
|
||||||
|
|
||||||
|
async def get_index_trade_stats_multi(
|
||||||
|
self,
|
||||||
|
index: str,
|
||||||
|
dates: Sequence[int | None],
|
||||||
|
*,
|
||||||
|
concurrency: int = 24,
|
||||||
|
batch_size: int = 50,
|
||||||
|
skip_errors: bool = True,
|
||||||
|
on_progress: Callable[[int, int], None] | None = None,
|
||||||
|
) -> list[IndexTradeStats]:
|
||||||
|
"""多日批量版 :meth:`get_index_trade_stats`(成分只取一次)。
|
||||||
|
|
||||||
|
用于「近 N 日同时点累计平均笔数」这类基准曲线:一次拿到 N 天数据,
|
||||||
|
避免每天重复请求成分表。
|
||||||
|
|
||||||
|
用法::
|
||||||
|
|
||||||
|
# 最近 5 个交易日的笔数曲线(日期自己给,节假日请自行跳过)
|
||||||
|
stats = await td.get_index_trade_stats_multi(
|
||||||
|
"000300", [20261016, 20261019, 20261020, 20261021, 20261022])
|
||||||
|
"""
|
||||||
|
market, code = self._index_pair(index)
|
||||||
|
members = [m.symbol for m in await self.get_index_members(code)]
|
||||||
|
out: list[IndexTradeStats] = []
|
||||||
|
for d in dates:
|
||||||
|
out.append(await self.get_index_trade_stats(
|
||||||
|
code, d, concurrency=concurrency, batch_size=batch_size,
|
||||||
|
members=members, skip_errors=skip_errors,
|
||||||
|
on_progress=on_progress))
|
||||||
|
return out
|
||||||
|
|
||||||
# ================================================================== #
|
# ================================================================== #
|
||||||
# 八、服务器文件(原始字节 + 解析)
|
# 八、服务器文件(原始字节 + 解析)
|
||||||
# ================================================================== #
|
# ================================================================== #
|
||||||
|
|||||||
@@ -8,6 +8,7 @@
|
|||||||
tdx transactions sz000001 --count 50 # 逐笔成交
|
tdx transactions sz000001 --count 50 # 逐笔成交
|
||||||
tdx minute sz000001 # 分时(今日/历史--date)
|
tdx minute sz000001 # 分时(今日/历史--date)
|
||||||
tdx index 000001 --count 30 # 指数K线
|
tdx index 000001 --count 30 # 指数K线
|
||||||
|
tdx index-stats 000300 --top 10 # 指数成分成交汇总(每分钟成交量/笔数)
|
||||||
tdx board list [--type 1] # 板块列表(0全部/1行业/2概念)
|
tdx board list [--type 1] # 板块列表(0全部/1行业/2概念)
|
||||||
tdx board members 881001 --count 20 # 板块成分
|
tdx board members 881001 --count 20 # 板块成分
|
||||||
tdx board ranking [--top 10] # 板块涨跌幅排行
|
tdx board ranking [--top 10] # 板块涨跌幅排行
|
||||||
@@ -204,6 +205,43 @@ async def cmd_board(td: TdxData, args) -> None:
|
|||||||
print(" %-12s %+8.2f%%" % (r.get("name", ""), r.get("change_pct", 0)))
|
print(" %-12s %+8.2f%%" % (r.get("name", ""), r.get("change_pct", 0)))
|
||||||
|
|
||||||
|
|
||||||
|
async def cmd_index_stats(td: TdxData, index: str, date: int | None,
|
||||||
|
show_all: bool, top: int, concurrency: int) -> None:
|
||||||
|
"""指数成分成交汇总:每分钟成交量 / 笔数(笔数=成分股逐笔累加)。"""
|
||||||
|
total = 0
|
||||||
|
|
||||||
|
def on_progress(done: int, all_: int) -> None:
|
||||||
|
nonlocal total
|
||||||
|
total = all_
|
||||||
|
print("\r 拉取成分逐笔 %d/%d ..." % (done, all_), end="", file=sys.stderr,
|
||||||
|
flush=True)
|
||||||
|
|
||||||
|
st = await td.get_index_trade_stats(
|
||||||
|
index, date, concurrency=concurrency, on_progress=on_progress)
|
||||||
|
if total:
|
||||||
|
print("\r" + " " * 40 + "\r", end="", file=sys.stderr, flush=True)
|
||||||
|
|
||||||
|
print("%s 成分成交汇总" % st.index)
|
||||||
|
print(" 成分 %d 只 / 有成交 %d 只 / 有笔数 %d 只%s" % (
|
||||||
|
st.member_count, st.active_members, st.trade_count_members,
|
||||||
|
"" if st.trade_count_available else " ★该指数无笔数口径"))
|
||||||
|
print(" 成交量 %s 手 成交笔数 %s 笔 成交额 %.2f 亿 耗时 %.1fs" % (
|
||||||
|
_fmt_num(st.total_volume),
|
||||||
|
_fmt_num(st.total_trade_count) if st.trade_count_available else "-",
|
||||||
|
st.total_amount / 1e8, st.elapsed))
|
||||||
|
if st.failed_members:
|
||||||
|
print(" (%d 只成分逐笔拉取失败,已跳过)" % st.failed_members)
|
||||||
|
print()
|
||||||
|
print(" 时间 累计成交量 累计笔数")
|
||||||
|
rows = st.cumulative()
|
||||||
|
if not show_all:
|
||||||
|
rows = rows[-top:]
|
||||||
|
for m in rows:
|
||||||
|
print(" %-6s %12s %12s" % (
|
||||||
|
m.time, _fmt_num(m.volume),
|
||||||
|
_fmt_num(m.trade_count) if st.trade_count_available else "-"))
|
||||||
|
|
||||||
|
|
||||||
async def cmd_info(td: TdxData, sym: str) -> None:
|
async def cmd_info(td: TdxData, sym: str) -> None:
|
||||||
s = parse_symbol(sym)
|
s = parse_symbol(sym)
|
||||||
info = await td.get_symbol_info(Market(s.market), s.code)
|
info = await td.get_symbol_info(Market(s.market), s.code)
|
||||||
@@ -390,6 +428,17 @@ def main() -> None:
|
|||||||
ix.set_defaults(func=lambda td, a: cmd_kline(
|
ix.set_defaults(func=lambda td, a: cmd_kline(
|
||||||
td, a.code, KlinePeriod.DAY, a.count, 0, Adjust.NONE))
|
td, a.code, KlinePeriod.DAY, a.count, 0, Adjust.NONE))
|
||||||
|
|
||||||
|
ist = add("index-stats", "指数成分成交汇总(每分钟成交量/笔数)")
|
||||||
|
ist.add_argument("index", help="指数代码,如 000300 / 000001 / sh000016")
|
||||||
|
ist.add_argument("--date", type=str, default=None, help="YYYYMMDD,缺省=当日")
|
||||||
|
ist.add_argument("--top", type=int, default=10, help="只显示最后 N 分钟(默认10)")
|
||||||
|
ist.add_argument("--all", dest="show_all", action="store_true",
|
||||||
|
help="显示全部分钟")
|
||||||
|
ist.add_argument("--concurrency", type=int, default=24, help="逐笔并发数")
|
||||||
|
ist.set_defaults(func=lambda td, a: cmd_index_stats(
|
||||||
|
td, a.index, int(a.date) if a.date else None,
|
||||||
|
a.show_all, a.top, a.concurrency))
|
||||||
|
|
||||||
b = add("board", "板块")
|
b = add("board", "板块")
|
||||||
b.add_argument("action", choices=["list", "members", "ranking", "change"])
|
b.add_argument("action", choices=["list", "members", "ranking", "change"])
|
||||||
b.add_argument("symbol", nargs="?", default="")
|
b.add_argument("symbol", nargs="?", default="")
|
||||||
|
|||||||
@@ -262,7 +262,11 @@ class TransactionRecord:
|
|||||||
|
|
||||||
@dataclass(slots=True)
|
@dataclass(slots=True)
|
||||||
class MinuteBar:
|
class MinuteBar:
|
||||||
"""分时(跨市场统一模型;A股 MAC 0x122D / 扩展市场 0x122D 同构)"""
|
"""分时(跨市场统一模型;A股 MAC 0x122D / 扩展市场 0x122D 同构)。
|
||||||
|
|
||||||
|
``hour``/``minute`` 是**真实时钟**(``0 <= hour < 24``),跨日的分时请用
|
||||||
|
:attr:`datetime` / :attr:`trading_date`,不要自己拿 ``hour`` 拼日期。
|
||||||
|
"""
|
||||||
price: float
|
price: float
|
||||||
vol: float
|
vol: float
|
||||||
avg_price: float = 0.0 # 均价(协议第2字段)
|
avg_price: float = 0.0 # 均价(协议第2字段)
|
||||||
@@ -273,6 +277,50 @@ class MinuteBar:
|
|||||||
code: str = ""
|
code: str = ""
|
||||||
date: int = 0 # YYYYMMDD
|
date: int = 0 # YYYYMMDD
|
||||||
fields: dict[str, float] = field(default_factory=dict) # 扩展市场额外字段(动量等)
|
fields: dict[str, float] = field(default_factory=dict) # 扩展市场额外字段(动量等)
|
||||||
|
day_offset: int = 0 # 相对 date 的自然日偏移:美股凌晨段为 +1
|
||||||
|
|
||||||
|
def __post_init__(self) -> None:
|
||||||
|
"""美股(EX 通道)分时时间归一化。
|
||||||
|
|
||||||
|
EX 分时的 ``minutes`` 字段是**自开盘起的连续分钟数**,不是「当天的分钟数」:
|
||||||
|
实测 ``usAAPL`` 返回 ``21:30`` ~ ``27:59``(390 条里 240 条的 ``hour`` ≥ 24)。
|
||||||
|
``27:59`` 实际是次日 ``03:59`` 北京时间。直接 ``minutes // 60`` 会得到
|
||||||
|
``hour=27`` 这种**非法时钟值**,任何 ``datetime(...)`` 拼接都会抛异常,
|
||||||
|
按小时聚合也会把凌晨段错误地归到同一天。
|
||||||
|
|
||||||
|
这里拆成「真实小时 + 日偏移」:``hour = (minutes // 60) % 24``,
|
||||||
|
``day_offset = (minutes // 60) // 24``。A股/港股/期货的 ``hour`` 本就合法
|
||||||
|
(A股合成时间 09:30~15:00、港股 09:30~16:00),**零行为变化**。
|
||||||
|
"""
|
||||||
|
if self.hour >= 24:
|
||||||
|
self.day_offset += self.hour // 24
|
||||||
|
self.hour %= 24
|
||||||
|
|
||||||
|
@property
|
||||||
|
def datetime(self) -> datetime:
|
||||||
|
"""该分时点的时刻(北京时间,已含跨日偏移)。"""
|
||||||
|
base = date(self.date // 10000, self.date // 100 % 100, self.date % 100)
|
||||||
|
d = base + timedelta(days=self.day_offset)
|
||||||
|
return datetime(d.year, d.month, d.day, self.hour, self.minute)
|
||||||
|
|
||||||
|
@property
|
||||||
|
def is_us(self) -> bool:
|
||||||
|
"""是否美股(EX 通道 ``US_STOCK``)。"""
|
||||||
|
return self.market in _US_EX_MARKETS
|
||||||
|
|
||||||
|
@property
|
||||||
|
def trading_date(self) -> str:
|
||||||
|
"""该分时点归属的交易日,``YYYY-MM-DD``。
|
||||||
|
|
||||||
|
美股:一个交易日横跨两个北京自然日(21:30~次日 04:00),凌晨段
|
||||||
|
(北京 hour < 12)归属**前一天**的美东交易日,与
|
||||||
|
:attr:`SecurityBar.trading_date` 口径一致。A股/港股/期货即 ``date`` 本身。
|
||||||
|
按交易日聚合分时(合成日线、日内统计)**必须用它**。
|
||||||
|
"""
|
||||||
|
base = self.datetime
|
||||||
|
if self.is_us and base.hour < _US_DAY_SPLIT_HOUR:
|
||||||
|
base -= timedelta(days=1)
|
||||||
|
return base.date().isoformat()
|
||||||
|
|
||||||
|
|
||||||
@dataclass(slots=True)
|
@dataclass(slots=True)
|
||||||
@@ -514,3 +562,82 @@ class FinancialRecord:
|
|||||||
market: int # 0=深 1=沪
|
market: int # 0=深 1=沪
|
||||||
report_date: int # 报告期 YYYYMMDD
|
report_date: int # 报告期 YYYYMMDD
|
||||||
fields: list[float] = field(default_factory=list)
|
fields: list[float] = field(default_factory=list)
|
||||||
|
|
||||||
|
|
||||||
|
# ===== 指数成分成交汇总(TdxData.get_index_trade_stats) =====
|
||||||
|
|
||||||
|
@dataclass(slots=True)
|
||||||
|
class MinuteTradeStat:
|
||||||
|
"""指数**全部成分股**汇总后的单分钟成交**增量**(不是累计)。
|
||||||
|
|
||||||
|
累计曲线用 :meth:`IndexTradeStats.cumulative`,不要自己再对同一字段反复累加。
|
||||||
|
"""
|
||||||
|
time: str # HH:MM
|
||||||
|
volume: float # 成交量合计(手,1 手 = 100 股)
|
||||||
|
trade_count: int # 成交笔数合计
|
||||||
|
amount: float # 成交额估算(元)= Σ(逐笔价 × 手 × 100)
|
||||||
|
|
||||||
|
|
||||||
|
@dataclass(slots=True)
|
||||||
|
class IndexTradeStats:
|
||||||
|
"""某个指数「全部成分股逐笔」汇总出的成交曲线。
|
||||||
|
|
||||||
|
为什么需要它:**通达信协议里指数本身没有成交笔数** —— 指数逐笔
|
||||||
|
(``sh000001``)的 ``trade_count`` 字段实际是「当时均价 × 1000」(见
|
||||||
|
:class:`TransactionRecord`)。所以「指数笔数」唯一可行的口径是
|
||||||
|
**把成分股的逐笔笔数累加**,本类即该口径。
|
||||||
|
|
||||||
|
口径校验(2026-10-22 实测):
|
||||||
|
|
||||||
|
- ``total_amount`` 是**成分股**合计。对「成分恰好就是指数覆盖面」的指数,它与
|
||||||
|
指数快照 ``amount`` **完全相等**:``sh000016`` 661.34 亿 = 661.34 亿、
|
||||||
|
``sh000300`` 2373.98 亿 = 2373.98 亿(比值 1.0000)。
|
||||||
|
- **但深市 ``399xxx`` 的指数成交额字段是「全板块」口径**,不是样本股:
|
||||||
|
实测 ``sz399001`` 快照 5081.40 亿 = 深市 2908 只股票合计 5081.01 亿、
|
||||||
|
``sz399006`` 2568.60 亿 = 创业板全部股票合计。所以拿深市指数 ``amount``
|
||||||
|
与 ``total_amount`` 比会得到 ~0.43 这类无意义比值——不是数据错,是口径不同。
|
||||||
|
- ``total_volume`` 单位是**手**(与指数逐笔的「百元」单位不同,勿混用)。
|
||||||
|
- 北交所成分(``bj`` 前缀)**没有笔数字段**(实测 ``trade_count`` 恒为 0):
|
||||||
|
此时 ``trade_count_members = 0``、:attr:`trade_count_available` 为 ``False``,
|
||||||
|
只有成交量可用。``sh``/``sz`` 成分均正常返回笔数。
|
||||||
|
|
||||||
|
``minutes`` 只包含**实际有成交**的分钟(午休 11:30~13:00 之间无桶),
|
||||||
|
画累计曲线时请按 ``time`` 自行补齐或直接连线。
|
||||||
|
"""
|
||||||
|
index: str # 规范化指数标的(如 sh000300)
|
||||||
|
date: int | None # None=当日;否则 YYYYMMDD
|
||||||
|
member_count: int # 指数成分数
|
||||||
|
active_members: int # 实际有逐笔成交的成分数(停牌/无成交不计)
|
||||||
|
trade_count_members: int # 返回了笔数的成分数(北交所为 0)
|
||||||
|
minutes: list[MinuteTradeStat] = field(default_factory=list)
|
||||||
|
total_volume: float = 0.0
|
||||||
|
total_trade_count: int = 0
|
||||||
|
total_amount: float = 0.0
|
||||||
|
failed_members: int = 0 # 逐笔拉取失败的成分数(skip_errors=True 时)
|
||||||
|
elapsed: float = 0.0 # 本次汇总耗时(秒)
|
||||||
|
|
||||||
|
@property
|
||||||
|
def trade_count_available(self) -> bool:
|
||||||
|
"""是否有笔数口径(北交所成分为 ``False``,此时 ``trade_count`` 全 0)。"""
|
||||||
|
return self.trade_count_members > 0
|
||||||
|
|
||||||
|
def cumulative(self) -> list[MinuteTradeStat]:
|
||||||
|
"""累计曲线:当日**截止该分钟**的累计量,时间升序。
|
||||||
|
|
||||||
|
这是「当日截止该分钟累计成交笔数」的直接数据源。
|
||||||
|
"""
|
||||||
|
out: list[MinuteTradeStat] = []
|
||||||
|
vol = 0.0
|
||||||
|
tc = 0
|
||||||
|
amt = 0.0
|
||||||
|
for m in self.minutes:
|
||||||
|
vol += m.volume
|
||||||
|
tc += m.trade_count
|
||||||
|
amt += m.amount
|
||||||
|
out.append(MinuteTradeStat(time=m.time, volume=vol,
|
||||||
|
trade_count=tc, amount=amt))
|
||||||
|
return out
|
||||||
|
|
||||||
|
def cum_trade_count(self) -> list[tuple[str, int]]:
|
||||||
|
"""``[(HH:MM, 累计笔数)]``,画「截止该分钟累计笔数」曲线用。"""
|
||||||
|
return [(m.time, m.trade_count) for m in self.cumulative()]
|
||||||
@@ -13,23 +13,46 @@ MAX_QUOTES = 80
|
|||||||
|
|
||||||
|
|
||||||
def price_decimal_digits(market: int, code: str) -> int:
|
def price_decimal_digits(market: int, code: str) -> int:
|
||||||
"""推断报价有效小数位:股票 2 位(/100)、ETF/基金 3 位(/1000)。
|
"""推断报价的小数位:**协议原始整数除以 ``10**该值``** 才是真实价格。
|
||||||
|
|
||||||
**沪市指数必须是 2 位**,此前把 ``000``/``8`` 一并判成 3 位,导致上证50、
|
四档:股票/指数/板块 ``2``、基金/ETF/B股 ``3``、债券/回购 ``4``。
|
||||||
上证指数、科创50 等**价格整体少 10 倍**。实证(2026-10-21):协议原始整数
|
|
||||||
与分时/逐笔价格比对,6 只沪市指数 + 2 只板块指数(880001/880305)全部是
|
该函数只是**离线推断**(``get_quotes`` 解析时拿不到证券清单)。真正的权威来源
|
||||||
``×10`` 错误,而 4 只沪市 ETF(510300/510500/588000/511010)、4 只沪市
|
是证券列表里协议自带的 ``SecurityInfo.decimal_point`` 字段——2026-10-22 拉取
|
||||||
股票、3 只科创板、3 只深市指数、深市股票/创业板/ETF 全部 ``×1`` 正确。
|
全量 51,982 只(沪 27,520 + 深 24,462)逐条比对,本函数与之一一吻合、**零例外**。
|
||||||
|
能对上是因为「三位前缀 → 小数位」在全量清单里没有任何一个前缀出现混合小数位。
|
||||||
|
|
||||||
|
历史修正(两次都是同一类错误:把某段前缀想当然地归错档):
|
||||||
|
|
||||||
|
- **2026-10-21**:沪市 ``000``/``8`` 被判成 3 位 → 上证指数/上证50/科创50/
|
||||||
|
板块指数**价格少 10 倍**(3713 读成 371)。它们实际是 2 位。
|
||||||
|
- **2026-10-22**:债券与回购全线偏差 —— 沪市 ``010/018/019/020/1xx/2xx``
|
||||||
|
与深市 ``10x/11x/12x/13x/14x/19x/50x/52x/56x`` 实际是 **4 位**,原先
|
||||||
|
沪市落到 2 位(**少 100 倍**)、深市落到 3 位(**少 10 倍**)。实测例:
|
||||||
|
``sh019742`` 110.228 读成 11022.8、``sh204001`` 1.385 读成 138.5、
|
||||||
|
``sz131810`` 1.38 读成 13.8。同时沪市 B 股 ``900xxx`` 是 3 位(原先 2 位)。
|
||||||
|
|
||||||
|
实证方法:拿 ``pre_close``(快照口径)除以上一交易日**日线收盘**(K线口径,
|
||||||
|
走 ``/1000`` 独立路径),比值即标度误差 —— 1.0 正确、10 少 10 倍、100 少 100 倍。
|
||||||
"""
|
"""
|
||||||
code = (code or "").strip().rstrip("\x00")
|
code = (code or "").strip().rstrip("\x00")
|
||||||
if market == 1: # SH
|
if market == 1: # 沪市
|
||||||
if code.startswith("5"):
|
if code.startswith("5"):
|
||||||
return 3 # ETF/基金
|
return 3 # 基金/ETF/LOF(50/51/52/56/58 系)
|
||||||
return 2 # 指数(000xxx/880xxx)、股票(60xxxx/688xxx)
|
if code.startswith("900"):
|
||||||
if market == 0: # SZ
|
return 3 # B股
|
||||||
if code.startswith("1"):
|
if code.startswith(("000", "36", "88", "99", "7", "6")):
|
||||||
return 3 # 基金/债券
|
return 2 # 指数/板块/优先股/配售/股票(60/601/603/605/688)
|
||||||
return 2 # 指数(399xxx)、股票(00xxxx/30xxxx)
|
return 4 # 债券(01x/1xx/2xx)、回购(204xxx)、ABS
|
||||||
|
if market == 0: # 深市
|
||||||
|
if code.startswith(("080", "158", "159", "160", "161", "162", "163",
|
||||||
|
"164", "165", "166", "167", "168", "169",
|
||||||
|
"180", "181")):
|
||||||
|
return 3 # 基金/ETF/LOF/配债
|
||||||
|
if code.startswith(("000", "001", "002", "003", "070", "140",
|
||||||
|
"200", "201", "300", "301", "302", "395", "399")):
|
||||||
|
return 2 # 股票/指数/B股/优先股
|
||||||
|
return 4 # 债券(10x-14x/19x)、回购(131xxx)、ABS(50x/52x/56x)
|
||||||
return 2
|
return 2
|
||||||
|
|
||||||
|
|
||||||
|
|||||||
@@ -457,6 +457,24 @@ async def test_boards(td):
|
|||||||
assert isinstance(chg, list)
|
assert isinstance(chg, list)
|
||||||
|
|
||||||
|
|
||||||
|
async def test_index_trade_stats(td):
|
||||||
|
"""指数成分成交汇总:成分 → 逐笔 → 每分钟量/笔数。"""
|
||||||
|
for name in ("get_index_members", "get_index_trade_stats",
|
||||||
|
"get_index_trade_stats_multi"):
|
||||||
|
_use(name)
|
||||||
|
mem = await td.get_index_members("000016")
|
||||||
|
assert isinstance(mem, list) and mem and mem[0].symbol
|
||||||
|
st = await td.get_index_trade_stats("000016")
|
||||||
|
assert st.index == "sh000016"
|
||||||
|
assert st.member_count >= 1
|
||||||
|
assert st.minutes and st.minutes[0].time == "09:30"
|
||||||
|
assert st.total_trade_count >= 1
|
||||||
|
assert st.cum_trade_count()[-1][1] == st.total_trade_count
|
||||||
|
multi = await td.get_index_trade_stats_multi("000016", [20261020, None])
|
||||||
|
assert len(multi) == 2
|
||||||
|
assert multi[0].date == 20261020 and multi[1].date is None
|
||||||
|
|
||||||
|
|
||||||
# --------------------------------------------------------------------------- #
|
# --------------------------------------------------------------------------- #
|
||||||
# 六、服务器文件
|
# 六、服务器文件
|
||||||
# --------------------------------------------------------------------------- #
|
# --------------------------------------------------------------------------- #
|
||||||
|
|||||||
@@ -0,0 +1,303 @@
|
|||||||
|
"""指数成分成交汇总(get_index_members / get_index_trade_stats)单元测试。
|
||||||
|
|
||||||
|
全部离线:成分与逐笔都用实例属性打补丁,不建立任何真实连接。
|
||||||
|
"""
|
||||||
|
from __future__ import annotations
|
||||||
|
|
||||||
|
from types import SimpleNamespace
|
||||||
|
|
||||||
|
import pytest
|
||||||
|
|
||||||
|
from pytdxdata import TdxData
|
||||||
|
from pytdxdata.models import (
|
||||||
|
IndexTradeStats,
|
||||||
|
MemberQuote,
|
||||||
|
MinuteTradeStat,
|
||||||
|
TransactionRecord,
|
||||||
|
)
|
||||||
|
|
||||||
|
# ---------- _index_pair:裸代码按指数语义解析 ----------
|
||||||
|
|
||||||
|
@pytest.mark.parametrize("text,expect", [
|
||||||
|
("000001", (1, "000001")), # 上证指数(不是平安银行!)
|
||||||
|
("sh000001", (1, "000001")),
|
||||||
|
("000300", (1, "000300")),
|
||||||
|
("sh000300", (1, "000300")),
|
||||||
|
("000905", (1, "000905")),
|
||||||
|
("999999", (1, "999999")),
|
||||||
|
("880305", (1, "880305")),
|
||||||
|
("399001", (0, "399001")), # 深证成指
|
||||||
|
("sz399006", (0, "399006")),
|
||||||
|
("899050", (2, "899050")), # 北证50
|
||||||
|
("bj899050", (2, "899050")),
|
||||||
|
("sh.000300", (1, "000300")), # 点号/冒号都容忍
|
||||||
|
("sz:399001", (0, "399001")),
|
||||||
|
])
|
||||||
|
def test_index_pair_parses(text, expect):
|
||||||
|
assert TdxData._index_pair(text) == expect
|
||||||
|
|
||||||
|
|
||||||
|
@pytest.mark.parametrize("text", [
|
||||||
|
"", "600000", "sz000001x", "sh12345", "12345", "abc", "zzz000001",
|
||||||
|
])
|
||||||
|
def test_index_pair_rejects(text):
|
||||||
|
with pytest.raises(ValueError):
|
||||||
|
TdxData._index_pair(text)
|
||||||
|
|
||||||
|
|
||||||
|
def test_index_pair_rejects_non_index_cn_code():
|
||||||
|
"""000001 之外的普通股票段(如 600xxx)不是指数,裸写必须报错。"""
|
||||||
|
with pytest.raises(ValueError):
|
||||||
|
TdxData._index_pair("600000")
|
||||||
|
|
||||||
|
|
||||||
|
# ---------- 模型:累计曲线 ----------
|
||||||
|
|
||||||
|
def test_index_trade_stats_cumulative():
|
||||||
|
st = IndexTradeStats(
|
||||||
|
index="sh000300", date=None, member_count=2, active_members=2,
|
||||||
|
trade_count_members=2,
|
||||||
|
minutes=[
|
||||||
|
MinuteTradeStat("09:30", 100.0, 10, 1000.0),
|
||||||
|
MinuteTradeStat("09:31", 200.0, 20, 2000.0),
|
||||||
|
MinuteTradeStat("09:32", 50.0, 5, 500.0),
|
||||||
|
],
|
||||||
|
total_volume=350.0, total_trade_count=35, total_amount=3500.0,
|
||||||
|
)
|
||||||
|
cum = st.cumulative()
|
||||||
|
assert [c.time for c in cum] == ["09:30", "09:31", "09:32"]
|
||||||
|
assert [c.volume for c in cum] == [100.0, 300.0, 350.0]
|
||||||
|
assert [c.trade_count for c in cum] == [10, 30, 35]
|
||||||
|
assert [c.amount for c in cum] == [1000.0, 3000.0, 3500.0]
|
||||||
|
# 累计必须是新对象,不污染原始增量
|
||||||
|
assert st.minutes[1].volume == 200.0
|
||||||
|
assert st.cum_trade_count() == [("09:30", 10), ("09:31", 30), ("09:32", 35)]
|
||||||
|
|
||||||
|
|
||||||
|
def test_index_trade_stats_trade_count_available():
|
||||||
|
ok = IndexTradeStats(index="sh000300", date=None, member_count=1,
|
||||||
|
active_members=1, trade_count_members=1)
|
||||||
|
assert ok.trade_count_available is True
|
||||||
|
bj = IndexTradeStats(index="bj899050", date=None, member_count=1,
|
||||||
|
active_members=1, trade_count_members=0)
|
||||||
|
assert bj.trade_count_available is False
|
||||||
|
|
||||||
|
|
||||||
|
# ---------- get_index_members:分页终止 ----------
|
||||||
|
|
||||||
|
def _member(market: int, code: str, name: str = "") -> MemberQuote:
|
||||||
|
return MemberQuote(market=market, code=code, name=name,
|
||||||
|
fields={"0x0": 9.9, "0x5": 1000.0, "0x7": 99000.0})
|
||||||
|
|
||||||
|
|
||||||
|
@pytest.mark.asyncio
|
||||||
|
async def test_get_index_members_pages_until_short_batch():
|
||||||
|
"""短页(不足 80 条)即视为最后一页,不再多发一次空请求。"""
|
||||||
|
calls: list[int] = []
|
||||||
|
|
||||||
|
async def fake_exec(cmd):
|
||||||
|
calls.append(cmd.start)
|
||||||
|
if cmd.start == 0:
|
||||||
|
return [_member(1, "600000"), _member(1, "600004")]
|
||||||
|
return []
|
||||||
|
|
||||||
|
td = TdxData(cache_dir=None)
|
||||||
|
td._execute_mac = fake_exec
|
||||||
|
members = await td.get_index_members("000016")
|
||||||
|
assert [m.symbol for m in members] == ["sh600000", "sh600004"]
|
||||||
|
assert calls == [0]
|
||||||
|
|
||||||
|
|
||||||
|
@pytest.mark.asyncio
|
||||||
|
async def test_get_index_members_continues_on_full_page():
|
||||||
|
"""满页(80 条)后继续翻页,直到空页。"""
|
||||||
|
calls: list[int] = []
|
||||||
|
|
||||||
|
async def fake_exec(cmd):
|
||||||
|
calls.append(cmd.start)
|
||||||
|
if cmd.start == 0:
|
||||||
|
return [_member(1, "600000")] * cmd.page_size
|
||||||
|
return []
|
||||||
|
|
||||||
|
td = TdxData(cache_dir=None)
|
||||||
|
td._execute_mac = fake_exec
|
||||||
|
members = await td.get_index_members("000016")
|
||||||
|
assert len(members) == 80
|
||||||
|
assert calls == [0, 80]
|
||||||
|
|
||||||
|
|
||||||
|
@pytest.mark.asyncio
|
||||||
|
async def test_get_index_members_respects_count_cap():
|
||||||
|
async def fake_exec(cmd):
|
||||||
|
# 永远返回满页,验证 count 能截断结果
|
||||||
|
return [_member(1, "600000")] * cmd.page_size
|
||||||
|
|
||||||
|
td = TdxData(cache_dir=None)
|
||||||
|
td._execute_mac = fake_exec
|
||||||
|
members = await td.get_index_members("000016", count=5)
|
||||||
|
assert len(members) == 5
|
||||||
|
|
||||||
|
|
||||||
|
# ---------- get_index_trade_stats:按分钟聚合 ----------
|
||||||
|
|
||||||
|
def _tick(sym: str, hhmm: str, price: float, vol: float, tc: int) -> TransactionRecord:
|
||||||
|
market, code = (1, sym[2:]) if sym.startswith("sh") else (0, sym[2:])
|
||||||
|
return TransactionRecord(market=market, code=code, time=f"{hhmm}:00",
|
||||||
|
price=price, vol=vol, trade_count=tc, bs_flag=1,
|
||||||
|
symbol=sym)
|
||||||
|
|
||||||
|
|
||||||
|
@pytest.mark.asyncio
|
||||||
|
async def test_get_index_trade_stats_aggregates_minutes():
|
||||||
|
async def fake_ticks(symbols, date=None, concurrency=6, **kw):
|
||||||
|
out = []
|
||||||
|
for s in symbols:
|
||||||
|
if s == "sh600000":
|
||||||
|
out += [_tick(s, "09:30", 10.0, 100.0, 10),
|
||||||
|
_tick(s, "09:30", 10.0, 50.0, 5),
|
||||||
|
_tick(s, "09:31", 11.0, 200.0, 20)]
|
||||||
|
else:
|
||||||
|
out += [_tick(s, "09:30", 5.0, 1000.0, 7)]
|
||||||
|
return out
|
||||||
|
|
||||||
|
td = TdxData(cache_dir=None)
|
||||||
|
td.get_ticks = fake_ticks
|
||||||
|
st = await td.get_index_trade_stats(
|
||||||
|
"000016", members=["sh600000", "sz000001"])
|
||||||
|
|
||||||
|
assert st.index == "sh000016"
|
||||||
|
assert st.member_count == 2
|
||||||
|
assert st.active_members == 2
|
||||||
|
assert st.trade_count_members == 2
|
||||||
|
assert st.trade_count_available is True
|
||||||
|
assert st.failed_members == 0
|
||||||
|
|
||||||
|
m = {x.time: x for x in st.minutes}
|
||||||
|
assert sorted(m) == ["09:30", "09:31"]
|
||||||
|
assert m["09:30"].volume == 1150.0 # 100 + 50 + 1000
|
||||||
|
assert m["09:30"].trade_count == 22 # 10 + 5 + 7
|
||||||
|
assert m["09:31"].volume == 200.0
|
||||||
|
# amount = Σ(price × 手 × 100)
|
||||||
|
assert m["09:30"].amount == pytest.approx(
|
||||||
|
10.0 * 100 * 100 + 10.0 * 50 * 100 + 5.0 * 1000 * 100)
|
||||||
|
assert st.total_volume == 1350.0
|
||||||
|
assert st.total_trade_count == 42
|
||||||
|
assert st.total_amount == pytest.approx(
|
||||||
|
10.0 * 100 * 100 + 10.0 * 50 * 100 + 5.0 * 1000 * 100 + 11.0 * 200 * 100)
|
||||||
|
assert st.cum_trade_count() == [("09:30", 22), ("09:31", 42)]
|
||||||
|
|
||||||
|
|
||||||
|
@pytest.mark.asyncio
|
||||||
|
async def test_get_index_trade_stats_bj_has_no_trade_count():
|
||||||
|
"""北交所成分返回 trade_count=0 → trade_count_available False。"""
|
||||||
|
async def fake_ticks(symbols, date=None, concurrency=6, **kw):
|
||||||
|
return [_tick("sh600000", "09:30", 10.0, 100.0, 0)]
|
||||||
|
|
||||||
|
td = TdxData(cache_dir=None)
|
||||||
|
td.get_ticks = fake_ticks
|
||||||
|
st = await td.get_index_trade_stats("899050", members=["sh600000"])
|
||||||
|
assert st.trade_count_members == 0
|
||||||
|
assert st.trade_count_available is False
|
||||||
|
assert st.total_volume == 100.0
|
||||||
|
|
||||||
|
|
||||||
|
@pytest.mark.asyncio
|
||||||
|
async def test_get_index_trade_stats_retries_per_symbol_on_batch_failure():
|
||||||
|
"""整批失败 → 退化为逐只重试,只有真正失败的计入 failed_members。"""
|
||||||
|
async def fake_ticks(symbols, date=None, concurrency=6, **kw):
|
||||||
|
single = isinstance(symbols, str)
|
||||||
|
if not single:
|
||||||
|
raise ConnectionError("批量炸了")
|
||||||
|
if symbols == "sh600004":
|
||||||
|
raise ConnectionError("这只也炸了")
|
||||||
|
return [_tick(symbols, "09:30", 10.0, 100.0, 3)]
|
||||||
|
|
||||||
|
td = TdxData(cache_dir=None)
|
||||||
|
td.get_ticks = fake_ticks
|
||||||
|
st = await td.get_index_trade_stats(
|
||||||
|
"000016", members=["sh600000", "sh600004", "sh600006"])
|
||||||
|
assert st.failed_members == 1
|
||||||
|
assert st.active_members == 2
|
||||||
|
assert st.total_trade_count == 6
|
||||||
|
|
||||||
|
|
||||||
|
@pytest.mark.asyncio
|
||||||
|
async def test_get_index_trade_stats_raises_when_not_skip_errors():
|
||||||
|
async def fake_ticks(symbols, date=None, concurrency=6, **kw):
|
||||||
|
raise ConnectionError("boom")
|
||||||
|
|
||||||
|
td = TdxData(cache_dir=None)
|
||||||
|
td.get_ticks = fake_ticks
|
||||||
|
with pytest.raises(ConnectionError):
|
||||||
|
await td.get_index_trade_stats(
|
||||||
|
"000016", members=["sh600000"], skip_errors=False)
|
||||||
|
|
||||||
|
|
||||||
|
@pytest.mark.asyncio
|
||||||
|
async def test_get_index_trade_stats_progress_callback():
|
||||||
|
seen: list[tuple[int, int]] = []
|
||||||
|
|
||||||
|
async def fake_ticks(symbols, date=None, concurrency=6, **kw):
|
||||||
|
return [_tick(s, "09:30", 10.0, 1.0, 1) for s in symbols]
|
||||||
|
|
||||||
|
td = TdxData(cache_dir=None)
|
||||||
|
td.get_ticks = fake_ticks
|
||||||
|
st = await td.get_index_trade_stats(
|
||||||
|
"000016", members=[f"sh60000{i}" for i in range(5)],
|
||||||
|
batch_size=2, on_progress=lambda d, t: seen.append((d, t)))
|
||||||
|
assert seen == [(2, 5), (4, 5), (5, 5)]
|
||||||
|
assert st.member_count == 5
|
||||||
|
|
||||||
|
|
||||||
|
@pytest.mark.asyncio
|
||||||
|
async def test_get_index_trade_stats_fetches_members_when_absent():
|
||||||
|
asked: list[str] = []
|
||||||
|
|
||||||
|
async def fake_members(code):
|
||||||
|
asked.append(code)
|
||||||
|
return [SimpleNamespace(symbol="sh600000")]
|
||||||
|
|
||||||
|
async def fake_ticks(symbols, date=None, concurrency=6, **kw):
|
||||||
|
return [_tick(s, "09:30", 10.0, 1.0, 1) for s in symbols]
|
||||||
|
|
||||||
|
td = TdxData(cache_dir=None)
|
||||||
|
td.get_index_members = fake_members
|
||||||
|
td.get_ticks = fake_ticks
|
||||||
|
await td.get_index_trade_stats("sh000300")
|
||||||
|
assert asked == ["000300"]
|
||||||
|
|
||||||
|
|
||||||
|
@pytest.mark.asyncio
|
||||||
|
async def test_get_index_trade_stats_rejects_empty_members():
|
||||||
|
td = TdxData(cache_dir=None)
|
||||||
|
|
||||||
|
async def fake_members(code):
|
||||||
|
return []
|
||||||
|
|
||||||
|
td.get_index_members = fake_members
|
||||||
|
with pytest.raises(ValueError, match="没有取到成分股"):
|
||||||
|
await td.get_index_trade_stats("000300")
|
||||||
|
|
||||||
|
|
||||||
|
@pytest.mark.asyncio
|
||||||
|
async def test_get_index_trade_stats_multi_reuses_members():
|
||||||
|
member_calls: list[str] = []
|
||||||
|
tick_dates: list[int | None] = []
|
||||||
|
|
||||||
|
async def fake_members(code):
|
||||||
|
member_calls.append(code)
|
||||||
|
return [SimpleNamespace(symbol="sh600000")]
|
||||||
|
|
||||||
|
async def fake_ticks(symbols, date=None, concurrency=6, **kw):
|
||||||
|
tick_dates.append(date)
|
||||||
|
return [_tick(s, "09:30", 10.0, 1.0, 2) for s in symbols]
|
||||||
|
|
||||||
|
td = TdxData(cache_dir=None)
|
||||||
|
td.get_index_members = fake_members
|
||||||
|
td.get_ticks = fake_ticks
|
||||||
|
out = await td.get_index_trade_stats_multi(
|
||||||
|
"000300", [20261020, 20261021, None])
|
||||||
|
assert len(out) == 3
|
||||||
|
assert member_calls == ["000300"] # 成分只取一次
|
||||||
|
assert tick_dates == [20261020, 20261021, None]
|
||||||
|
assert all(s.total_trade_count == 2 for s in out)
|
||||||
+110
-1
@@ -337,7 +337,37 @@ def test_finish_bars_sorts_us_session_chronologically():
|
|||||||
(0, "000001", 2), # 平安银行(不是指数!)
|
(0, "000001", 2), # 平安银行(不是指数!)
|
||||||
(0, "300750", 2),
|
(0, "300750", 2),
|
||||||
(0, "159915", 3), # 深市 ETF
|
(0, "159915", 3), # 深市 ETF
|
||||||
(0, "128035", 3), # 深市转债
|
# 债券/回购:2026-10-22 修正(沪市曾少 100 倍、深市曾少 10 倍)
|
||||||
|
(1, "019742", 4), # 沪市国债 24特国01(110.228,曾读成 11022.8)
|
||||||
|
(1, "019547", 4), # 沪市国债
|
||||||
|
(1, "010706", 4), # 沪市老国债 07国债06
|
||||||
|
(1, "018014", 4), # 沪市 018 段
|
||||||
|
(1, "020812", 4), # 沪市贴债
|
||||||
|
(1, "110077", 4), # 沪市可转债 洪城转债(199.279,曾读成 19927.9)
|
||||||
|
(1, "113616", 4), # 沪市可转债 韦尔转债
|
||||||
|
(1, "118031", 4), # 沪市可转债
|
||||||
|
(1, "111000", 4), # 沪市 111 段
|
||||||
|
(1, "132024", 4), # 沪市可交换债
|
||||||
|
(1, "184832", 4), # 沪市 184 段
|
||||||
|
(1, "204001", 4), # 沪市国债逆回购 GC001(1.385,曾读成 138.5)
|
||||||
|
(1, "204028", 4),
|
||||||
|
(0, "123107", 4), # 深市可转债 温氏转债(112.189,曾读成 1121.89)
|
||||||
|
(0, "128136", 4), # 深市可转债 立讯转债
|
||||||
|
(0, "127025", 4), # 深市可转债
|
||||||
|
(0, "131810", 4), # 深市国债逆回购 R-001(1.38,曾读成 13.8)
|
||||||
|
(0, "148004", 4), # 深市 148 段
|
||||||
|
(0, "112001", 4), # 深市公司债
|
||||||
|
(0, "101917", 4), # 深市国债
|
||||||
|
(0, "520261", 4), # 深市 520 段
|
||||||
|
# 沪市 B 股:3 位(2026-10-22 修正,曾按 2 位少 10 倍)
|
||||||
|
(1, "900901", 3), # 云赛B股
|
||||||
|
(1, "900948", 3),
|
||||||
|
# 深市 B 股:2 位(与沪市不同!)
|
||||||
|
(0, "200011", 2), # 深物业B
|
||||||
|
(0, "200488", 2),
|
||||||
|
# 边界:999xxx 是「指数」,不能被 9 开头当成 B 股
|
||||||
|
(1, "999999", 2), # 上证指数
|
||||||
|
(1, "999998", 2), # A股指数
|
||||||
])
|
])
|
||||||
def test_price_decimal_digits(market, code, expect):
|
def test_price_decimal_digits(market, code, expect):
|
||||||
assert price_decimal_digits(market, code) == expect
|
assert price_decimal_digits(market, code) == expect
|
||||||
@@ -345,3 +375,82 @@ def test_price_decimal_digits(market, code, expect):
|
|||||||
|
|
||||||
def test_price_decimal_digits_strips_nul():
|
def test_price_decimal_digits_strips_nul():
|
||||||
assert price_decimal_digits(1, "000016\x00\x00") == 2
|
assert price_decimal_digits(1, "000016\x00\x00") == 2
|
||||||
|
|
||||||
|
|
||||||
|
# ---------- 美股分时跨日:hour 必须是合法时钟值(回归:曾出现 hour=27) ----------
|
||||||
|
|
||||||
|
def _us_minutes(n: int = 390) -> list:
|
||||||
|
"""构造 usAAPL 的 390 条分时:从北京 21:30 起连续 n 分钟。"""
|
||||||
|
from pytdxdata.models import MinuteBar
|
||||||
|
|
||||||
|
start = 21 * 60 + 30
|
||||||
|
return [
|
||||||
|
MinuteBar(price=100.0 + i * 0.01, vol=1.0,
|
||||||
|
hour=(start + i) // 60, minute=(start + i) % 60,
|
||||||
|
market=74, code="AAPL", date=20261022)
|
||||||
|
for i in range(n)
|
||||||
|
]
|
||||||
|
|
||||||
|
|
||||||
|
def test_minute_bar_us_hour_is_valid_clock_value():
|
||||||
|
"""EX 分时的 minutes 是「自开盘起的连续分钟数」,直接 //60 会得到 hour=27。"""
|
||||||
|
bars = _us_minutes()
|
||||||
|
assert all(0 <= b.hour < 24 for b in bars)
|
||||||
|
# 归一化前的原始值确实越界(21:30 + 390min = 28:00)
|
||||||
|
assert max(b.hour + b.day_offset * 24 for b in bars) > 23
|
||||||
|
|
||||||
|
|
||||||
|
def test_minute_bar_us_day_offset_marks_next_day():
|
||||||
|
bars = _us_minutes()
|
||||||
|
assert bars[0].day_offset == 0 # 21:30 当晚
|
||||||
|
assert bars[-1].day_offset == 1 # 03:59 次日
|
||||||
|
assert (bars[-1].hour, bars[-1].minute) == (3, 59)
|
||||||
|
|
||||||
|
|
||||||
|
def test_minute_bar_us_datetime_is_monotonic_across_midnight():
|
||||||
|
bars = _us_minutes()
|
||||||
|
dts = [b.datetime for b in bars]
|
||||||
|
assert dts == sorted(dts)
|
||||||
|
assert dts[0] == datetime(2026, 10, 22, 21, 30)
|
||||||
|
assert dts[-1] == datetime(2026, 10, 23, 3, 59)
|
||||||
|
|
||||||
|
|
||||||
|
def test_minute_bar_us_trading_date_spans_two_calendar_days():
|
||||||
|
"""一个美股交易日的分时横跨两个北京自然日,trading_date 必须都是同一天。"""
|
||||||
|
bars = _us_minutes()
|
||||||
|
tds = {b.trading_date for b in bars}
|
||||||
|
assert tds == {"2026-10-22"}
|
||||||
|
# 而 datetime.date() 会给出两天 —— 这正是不能用它的原因
|
||||||
|
assert len({b.datetime.date() for b in bars}) == 2
|
||||||
|
|
||||||
|
|
||||||
|
def test_minute_bar_cn_unchanged():
|
||||||
|
"""A股/港股分时 hour 本就合法,归一化必须零行为变化。"""
|
||||||
|
from pytdxdata.models import MinuteBar
|
||||||
|
|
||||||
|
for hour, minute in ((9, 30), (11, 29), (13, 0), (14, 59)):
|
||||||
|
b = MinuteBar(price=10.0, vol=1.0, hour=hour, minute=minute,
|
||||||
|
market=1, code="600000", date=20261022)
|
||||||
|
assert (b.hour, b.minute) == (hour, minute)
|
||||||
|
assert b.day_offset == 0
|
||||||
|
assert b.datetime == datetime(2026, 10, 22, hour, minute)
|
||||||
|
assert b.trading_date == "2026-10-22"
|
||||||
|
|
||||||
|
|
||||||
|
def test_minute_bar_cn_datetime_monotonic_full_session():
|
||||||
|
"""A股全天 240 条(09:30~11:29 + 13:00~14:59)时间轴严格递增。"""
|
||||||
|
from pytdxdata.models import MinuteBar
|
||||||
|
|
||||||
|
bars = []
|
||||||
|
for idx in range(240):
|
||||||
|
if idx < 120:
|
||||||
|
hour, minute = 9 + (30 + idx) // 60, (30 + idx) % 60
|
||||||
|
else:
|
||||||
|
hour, minute = 13 + (idx - 120) // 60, (idx - 120) % 60
|
||||||
|
bars.append(MinuteBar(price=10.0, vol=1.0, hour=hour, minute=minute,
|
||||||
|
market=1, code="600000", date=20261022))
|
||||||
|
dts = [b.datetime for b in bars]
|
||||||
|
assert dts == sorted(dts)
|
||||||
|
assert dts[0] == datetime(2026, 10, 22, 9, 30)
|
||||||
|
assert dts[-1] == datetime(2026, 10, 22, 14, 59)
|
||||||
|
assert {b.trading_date for b in bars} == {"2026-10-22"}
|
||||||
Reference in new issue
Block a user