[ENH] Remove yfinance as a dependency and implement data_loader

pypfopt/data/__init__.py

@@ -0,0 +1,3 @@
+from .data_loader import load_stockdata, available_tickers, load_marketcaps
+
+__all__ = ["load_stockdata", "available_tickers", "load_marketcaps"]

pypfopt/data/data_loader.py

@@ -0,0 +1,47 @@
+import pandas as pd
+from importlib import resources
+
+
+def _load_raw_data(filename: str, **read_csv_kwargs):
+    with resources.files(__package__).joinpath(filename).open("r") as f:
+        return pd.read_csv(f, **read_csv_kwargs)
+
+
+def load_stockdata(tickers: list = None, start: str = None, end: str = None):
+
+    df = _load_raw_data("stock_prices.csv", parse_dates=["date"])
+
+    if start is not None:
+        df = df[df["date"] >= pd.to_datetime(start)]
+    if end is not None:
+        df = df[df["date"] <= pd.to_datetime(end)]
+
+    if tickers is not None:
+        cols = ["date"] + tickers
+        df = df[cols]
+
+    return df.set_index("date")
+
+
+def load_marketcaps(tickers: list = None):
+
+    df = _load_raw_data("market_caps.csv")
+
+    if tickers is not None:
+        available = set(df["ticker"])
+        invalid = set(tickers) - available
+        if invalid:
+            raise ValueError(f"Invalid tickers: {invalid}")
+
+        df = df[df["ticker"].isin(tickers)]
+
+    return dict(zip(df["ticker"], df["market_cap"]))
+
+
+def available_tickers():
+
+    df = _load_raw_data("stock_prices.csv", parse_dates=["date"])
+    cols = [c for c in df.columns if c != "date"]
+    cols.sort()
+
+    return cols

pypfopt/data/market_caps.csv

@@ -0,0 +1,34 @@
+ticker,market_cap
+AAPL,752207537100
+ACN,1901675041287
+AMD,1465327913913
+AMZN,1199323675973
+BAC,316257187682
+BLK,316209068070
+COST,120876806275
+CVS,1733021410820
+DIS,1204224448427
+DPZ,1417604792703
+F,46066066120
+GILD,1939970155063
+INTU,1665723068396
+JD,428616525803
+JPM,367740809578
+KO,370891997157
+LUV,611963274704
+MA,1051889081106
+MCD,866730312191
+MSFT,586002134695
+NAT,1225646524971
+NVDA,283290252000
+PBI,587828573827
+PFE,735891877370
+SBUX,914859618512
+SPY,1571426042979
+TGT,403349195405
+TM,1030897704635
+TSLA,1186867064879
+UL,97668573376
+UNH,1217051979543
+WMT,345195626756
+XOM,134777928005

pyproject.toml

@@ -62,7 +62,6 @@ all_extras = [
 dev = [
     "pytest>=9.0.0",
     "pytest-cov>=7.0.0",
-    "yfinance>=0.2.66",
 ]
 
 # notebook tests
@@ -87,6 +86,9 @@ requires = [
 [tool.setuptools.packages.find]
 exclude = ["example", "example.*", "tests", "tests.*"]
 
+[tool.setuptools.package-data]
+pypfopt = ["data/*.csv"]
+
 [tool.ruff]
 line-length = 88
 # Keep Ruff aligned with project target version