Files
sanguo_vnpy_v2/sanguo_factor/fundamental_schema.py
T
claude_dev 0f8a97a219
CI/CD / test (push) Successful in 24s
CI/CD / nas-deploy (push) Successful in 13s
CI/CD / nas-verify (push) Successful in 6s
feat(factor): P1-C 批研报共识 5+质押 4 因子 TDD(spec §20.13 三新域消费面) [nas]
R 研报共识族(PIT=publishDate,180 天窗两端 cum 口径同 E11 手法):
- res_rating_pos 窗内买入+增持占比(emRatingName 正面集={买入,增持,推荐,强推,强买}),
  覆盖门窗内研报<3 条→NaN 稀样本失真宁缺毋假
- res_coverage 窗内研报条数(首研报前 NaN=未覆盖,曾覆盖后窗空=0 值两语义区分)
- res_fwd_ep 窗内 mean 预测 EPS/predictThisYearEps(空值不算分母;当年口径,
  跨年跳变由截面同年份对齐吸收)
- res_target_upside 最新带目标价条目+180 天新鲜度门——indvAimPriceT null 条目
  不进 target 流→不遮蔽上一份带价研报(join_asof 保留 eff 算 datetime−eff≤180)
- res_rating_chg 窗内净上调次数(ratingChange='上调'−'下调';EM 空值已统一转 None)

Q 质押族(无质押=0 真实零,fill 语义与缺域 NaN 严格区分):
- pledge_ratio 周五快照 asof backward(akshare 中文列,质押比例=总股本口径%)
- pledge_ctrl 实控人在押 max 占所持——pledge_detail 月桶区间重建:生效开始=
  max(PF_START_DATE,NOTICE_DATE) 公告日才可见宁晚毋早;结束=UNFREEZE_DATE
  (EM 无解押公告日列,假设解押日≈公告日);IS_CONTROL_SHAREHOLDER=='1' 白捡
  直标免 F10 actual_controller join;断点稀疏流=每股在押集合变化时刻聚合
  max(PF_HOLD_RATIO),断点间值恒定 join_asof 填格子
- pledge_margin_dist=−pledge_warn_line/close(在押最高预估平仓线占现价,
  低=安全垫厚;无在押 fill 0→0 值=最优档天然参与截面)
- pledge_net_180 窗内解押−新增笔数(生效开始=−1/解押=+1 两事件 cum 两端差);
  MXID 跨月桶去重取最新桶状态(文件名升序 unique keep last)
- 区间重建 1997 全史留存→全窗可回测,非前向积累(与 S 族对照)

工程: conftest 扩三新域合成树(周窗文件+周五快照+月桶断点/解押/PIT notice
晚 start 用例);计数列聚合即 cast Int64(u32 环回坑 P2 实锤先例延续);
FEATURE_COLUMNS+9 列/_DOMAIN_COLS 扩;既有计数断言 80→89;新增 16 用例,
全套 1458 绿;设计档 §3 表同步 P1-C 行 [nas]

Co-Authored-By: Claude Code <noreply@anthropic.com>
2026-09-22 14:11:07 +08:00

154 lines
9.2 KiB
Python
Raw Blame History

This file contains ambiguous Unicode characters
This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.
# sanguo_factor/fundamental_schema.py
"""财务因子适配层常量/列契约(自 fundamental_adapter.py 拆出,纯结构重构).
口径红线与数据形态的完整说明见 fundamental_adapter.py 模块 docstring;
本模块只承载各数据域共享的列映射/清单常量,被读取层/特征层/PIT 层共同引用.
"""
from __future__ import annotations
DEFAULT_STATIC_DIR = "/volume1/stock/sanguo_vnpy_v2/data/static"
_SYM = "vt_symbol" # 时序分组键
# vt_symbol 后缀 → NAS 文件名后缀(600000.SSE → 600000.SH)
_VT_TO_FILE_SUFFIX = {"SSE": "SH", "SZSE": "SZ"}
# forecast 预告类型 → 有序分(§3.6 F04)
FORECAST_TYPE_SCORE: dict[str, int] = {
"预增": 3, "略增": 2, "扭亏": 2, "续盈": 1, "减亏": 1, "不确定": 0,
"略减": -1, "增亏": -2, "续亏": -2, "预减": -3, "首亏": -3,
}
# 累计口径列映射: 源列 → 短名(需单季化+TTM)
_CUM_MAP = {
"TOTAL_OPERATE_INCOME": "rev",
"OPERATE_COST": "cogs",
"PARENT_NETPROFIT": "np",
"DEDUCT_PARENT_NETPROFIT": "dnp",
"TOTAL_PROFIT": "tp",
"INVEST_INCOME": "invest_inc",
"FAIRVALUE_CHANGE_INCOME": "fv_inc",
"ASSET_IMPAIRMENT_LOSS": "asset_imp",
"CREDIT_IMPAIRMENT_LOSS": "credit_imp",
"NETCASH_OPERATE": "cfo",
"SALES_SERVICES": "sales_cash",
"ACCEPT_INVEST_CASH": "acc_inv_cash",
# ---- P1 批新增(income;NAS 实测 2026-09-08 列名核实)----
"RESEARCH_EXPENSE": "research", # 研发费用(2018Q3 起单列,早年 null 传播)
"SALE_EXPENSE": "sale_exp", # 销售费用
"FE_INTEREST_EXPENSE": "int_exp", # 财务费用-利息费用(披露稀疏,EBIT 组装层按 0)
"INCOME_TAX": "tax", # 所得税费用(τ 实际税率分子)
"BASIC_EPS": "eps", # 基本每股收益(累计口径,需单季化)
# ---- P1 批新增(cashflow 补充资料段 + 融资流)----
"FA_IR_DEPR": "fa_depr", # 固定资产折旧
"IA_AMORTIZE": "ia_amort", # 无形资产摊销
"LPE_AMORTIZE": "lpe_amort", # 长期待摊费用摊销
"USERIGHT_ASSET_AMORTIZE": "ua_amort", # 使用权资产折旧摊销(2019 起才有)
"CONSTRUCT_LONG_ASSET": "capex", # 购建长期资产现金(C17 投资率/D15 FCF)
"RECEIVE_LOAN_CASH": "recv_loan", # 取得借款现金(E03)
"ISSUE_BOND": "issue_bond", # 发行债券现金(E03)
"PAY_DEBT_CASH": "pay_debt", # 偿还债务现金(E03)
}
# 存量列(balance,时点值直接用;缺列 → null)
_BALANCE_COLS = ["TOTAL_ASSETS", "TOTAL_PARENT_EQUITY", "ACCOUNTS_RECE",
"OTHER_RECE", "GOODWILL", "SHARE_CAPITAL", "SHORT_LOAN",
"SHORT_FIN_PAYABLE", "NONCURRENT_LIAB_1YEAR", "LONG_LOAN",
"BOND_PAYABLE", "LEASE_LIAB",
# P1 新增存量列
"INVENTORY", "MONETARYFUNDS",
# P1-B 新增存量列(E08 CCC 分子: 应付账款,NAS 实测列名核实)
"ACCOUNTS_PAYABLE"]
# IBD 口径钉死含 LEASE_LIAB 版(survey §3.5 E 族注意点「定稿钉死」;2019 前该列
# 整体缺失按 0,与一年内到期非流动负债等组件同款处理)
_IBD_PARTS = ["SHORT_LOAN", "SHORT_FIN_PAYABLE", "NONCURRENT_LIAB_1YEAR",
"LONG_LOAN", "BOND_PAYABLE", "LEASE_LIAB"]
# DA 组装: cashflow 间接法补充资料四件,列/值缺失按 0(「列存在才加」;
# USERIGHT_ASSET_AMORTIZE 2019 前整列缺失不影响早年 DA)
_DA_PARTS = ["fa_depr", "ia_amort", "lpe_amort", "ua_amort"]
_DATE_COLS = ["REPORT_DATE", "NOTICE_DATE", "UPDATE_DATE"]
# 输出特征列(P0 32 + P1 35 因子及 2 变体的全部原料;契约由 test_fundamental_library
# 与 test_fundamental_p1_library 锁定)
FEATURE_COLUMNS: list[str] = [
# 报告期级比率(盈利能力 A / 盈利质量 B / 成长 C / 资本结构 E / 预期事件 F)
"roe_ttm", "roe_deduct_ttm", "roa_ttm", "gp_over_assets", "gross_margin",
"net_margin", "cfo_over_assets",
"tacc", "nonrec_ratio", "impairment_ratio", "invest_income_dep",
"receivables_anomaly", "sales_cash_ratio", "other_rece_ratio",
"rev_q_yoy", "np_q_yoy", "growth_scissors", "gm_delta", "roe_delta",
"asset_growth", "nsi", "ibd_ratio", "goodwill_ratio",
"sue_np", "sue_rev",
"forecast_type_score", "forecast_change_pct",
# 估值/资本行为因子的日频原料(表达式层 ÷ close×share_capital)
"np_ttm", "dnp_ttm", "cfo_ttm", "equity", "share_capital",
"acc_invest_cash_ttm",
# ---- P1 批新增(A 盈利 7)----
"roe_avg", "roa_pretax", "ebit_over_assets", "ebitda_margin", "roic",
"rd_intensity", "sale_expense_ratio",
# ---- P1 批新增(B 质量 8;B12 为哑变量的连续近似乘积)----
"inventory_anomaly", "cash_ibd_product", "vsig", "vsig_acc", "vsig_cfo",
"da_intensity", "gm_nm_scissors", "profit_streak",
# ---- P1 批新增(C 成长 8)----
"np_accel", "rev_accel", "nm_delta", "rev_cagr5", "np_cagr5",
"nwc_growth", "invest_growth", "equity_growth",
# ---- P1 批新增(D 估值 6 的原料;EV = close×share_capital + ev_ex_mv;
# forecast_np_annualized 走 forecast 事件流公告日 asof)----
"ebit_ttm", "ebitda_ttm", "fcf_ttm", "gp_ttm", "ev_ex_mv",
"forecast_np_annualized",
# ---- P1 批新增(E 资本 3)----
"debt_issue_ttm", "interest_cover", "goodwill_growth",
# ---- P1 批新增(F 预期 3 + SUE 严窗变体;forecast_beat 走独立兑现差事件流)----
"sue_eps", "disclosure_speed", "sue_np_strict", "forecast_beat",
# ---- P1-B 批新增(四新域 5 因子;契约由 test_fundamental_p1b_* 锁定)----
"ccc", # E08 三表自算现金转换周期
"send_total_12m", # E11 dividend 事件流(365 天滚动合计)
"gdhs_chg", # E12 股东户数相邻事件 Δln
"topholder_chg", # E13 十大流通占比相邻期 Δ(聚合产物)
"ep_vb", "bp_vb", # D14 长史 EP/BP(vb 日频域)
# ---- P2 批新增(七项目指标调研增量: D 估值 2 + E 股息 3;契约由
# test_fundamental_p2_* 锁定)----
"np_ttm_yoy", # D16 PEG 原料: TTM 净利同比(_yoy4 ttm 口径)
"rev_ttm", # D17 PS 原料: 营收 TTM
"peg_denom", # D16 PEG 分母 = np_ttm×g(守卫 np_ttm>0 且 g>1%)
"div_12m", # E14 股息率分子: 近365天现金分红/股(除息日锚)
"div_payout", # E15 派息率(np_ttm>0 守卫,pit 层后算)
"div_streak", # E16 连续现金分红年数(断年重计)
# ---- P1-C 批新增(spec §20.13 三新域: 研报共识 5 + 质押 4;契约由
# test_fundamental_p1c_* 锁定)----
"res_coverage", # R01 原料: 180 天窗研报条数(publishDate 锚两端 cum)
"res_rating_pos", # R02 原料: 窗内买入+增持占比(门 n≥3)
"res_fwd_eps", # R03 原料: 窗内 mean 预测 EPS(空值不算分母)
"res_target_price", # R04 原料: 最新可见目标价(180 天新鲜度门)
"res_rating_chg", # R05 原料: 窗内净上调次数(上调−下调)
"pledge_ratio", # Q01 原料: 总股本质押比例%(周五快照 asof,无=0)
"pledge_ctrl", # Q02 原料: 实控人在押 max 占所持%(区间重建)
"pledge_warn_line", # Q03 原料: 在押最高预估平仓线价(无在押=0)
"pledge_net_180", # Q04 原料: 180 天窗解押−新增笔数(无=0)
]
# 金融股置 NaN 的特征(盈利质量 B 族 + 成长 C 族 + 费用类,§7 红线 5;
# survey A 族注意点: A14/A16 金融无三费结构 → rd/sale_expense_ratio 同剔)
_FIN_NULL_COLS = ["tacc", "nonrec_ratio", "impairment_ratio",
"invest_income_dep", "receivables_anomaly",
"sales_cash_ratio", "other_rece_ratio",
"rev_q_yoy", "np_q_yoy", "growth_scissors", "gm_delta",
"roe_delta", "asset_growth",
"rd_intensity", "sale_expense_ratio",
"inventory_anomaly", "cash_ibd_product", "vsig", "vsig_acc",
"vsig_cfo", "da_intensity", "gm_nm_scissors", "profit_streak",
"np_accel", "rev_accel", "nm_delta", "rev_cagr5", "np_cagr5",
"nwc_growth", "invest_growth", "equity_growth",
# P1-B: E08 CCC 属营运效率,金融股无 CCC 概念(§7 红线扩展)
"ccc"]
# forecast 事件流日频列(公告日 asof): P0 两列 + P1 年化预告净利(D13)
_FORECAST_COLS = ("forecast_type_score", "forecast_change_pct",
"forecast_np_annualized")
# 预告兑现差独立事件流(F07 前视红线: 锚 = max(实际披露日, 预告公告日))
_BEAT_COLS = ("forecast_beat",)
# P1-B 四新域列(非报告期特征,由各域事件/日频流产出;报表加载须排除)
# P2 股息三列(dividend 域除息日锚事件流,pit 层产出)同属域列;
# P1-C 研报共识 5 + 质押 4(三新域事件/快照流)同属域列
_DOMAIN_COLS = ("send_total_12m", "gdhs_chg", "topholder_chg", "ep_vb", "bp_vb",
"div_12m", "div_payout", "div_streak",
"res_coverage", "res_rating_pos", "res_fwd_eps",
"res_target_price", "res_rating_chg",
"pledge_ratio", "pledge_ctrl", "pledge_warn_line", "pledge_net_180")