liuyu/qlib
1
0
Fork 0
mirror of https://github.com/microsoft/qlib.git synced 2026-06-06 05:51:17 +08:00
Code Activity

fix: replace deprecated pandas fillna(method=) with ffill()/bfill() (#1987)

Browse Source 
* fix: replace deprecated pandas fillna(method=) with ffill()/bfill()

  Replace deprecated fillna(method="ffill"/"bfill") calls with modern
  pandas ffill() and bfill() methods to fix FutureWarnings in pandas 2.x.

  Also includes black formatting fixes for compliance.

  This addresses the pandas deprecation warnings portion of issue #1981.
  Other issues (date parsing, type conversion, timezone handling) will be
  addressed in separate commits.

  Fixes:
  - Yahoo collector: 2 instances in calc_change() and adjusted_price()
  - BaoStock collector: 1 instance in calc_change()
  - Core utils: resam.py fillna operations
  - Backtest: profit_attribution.py stock data processing
  - High-freq ops: FFillNan and BFillNan operators
  - Position analysis: parse_position.py weight processing

  Partially addresses GitHub issue #1981

* lint with black

* lint with black

* limit minimum version of pandas

* limit minimum version of pandas

---------

Co-authored-by: Linlang <Lv.Linlang@hotmail.com>
This commit is contained in:
Alaa Kaddour
2025-08-19 09:00:29 +01:00
committed by GitHub
parent 2d05a705e3
commit 7095e755fa
9 changed files with 13 additions and 12 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
scripts/data_collector/baostock_5min/collector.py
View File

@@ -172,7 +172,7 @@ class BaostockNormalizeHS3005min(BaseNormalize):
@staticmethod
def calc_change(df: pd.DataFrame, last_close: float) -> pd.Series:
df = df.copy()
_tmp_series = df["close"].fillna(method="ffill")
_tmp_series = df["close"].ffill()
_tmp_shift_series = _tmp_series.shift(1)
if last_close is not None:
_tmp_shift_series.iloc[0] = float(last_close)

4
scripts/data_collector/yahoo/collector.py
View File

@@ -371,7 +371,7 @@ class YahooNormalize(BaseNormalize):
@staticmethod
def calc_change(df: pd.DataFrame, last_close: float) -> pd.Series:
df = df.copy()
_tmp_series = df["close"].fillna(method="ffill")
_tmp_series = df["close"].ffill()
_tmp_shift_series = _tmp_series.shift(1)
if last_close is not None:
_tmp_shift_series.iloc[0] = float(last_close)
@@ -459,7 +459,7 @@ class YahooNormalize1d(YahooNormalize, ABC):
df.set_index(self._date_field_name, inplace=True)
if "adjclose" in df:
df["factor"] = df["adjclose"] / df["close"]
df["factor"] = df["factor"].fillna(method="ffill")
df["factor"] = df["factor"].ffill()
else:
df["factor"] = 1
for _col in self.COLUMNS: