fix cache implementation

Author: dakixr

src/isbtchot/model.py

@@ -6,21 +6,25 @@
 from isbtchot.schemas.args import TypeTime
 from sklearn.linear_model import LinearRegression
 
-USE_CACHE = True
-
 CACHE_BTC_PATH = isbtchot.root_path / "cache" / "btc.csv"
 BTC_API = (
     "https://min-api.cryptocompare.com/data/v2/histoday?fsym=BTC&tsym=USD&allData=true"
 )
 
-
-def btc_historical_daily() -> pd.DataFrame:
-    # Ensure the directory exists
-    CACHE_BTC_PATH.parent.mkdir(parents=True, exist_ok=True)
+def btc_historical_daily(options=None) -> pd.DataFrame:
+    
+    if not options:
+        options = {}
+        
+    use_cache = options.get("disable_cache") is None
+    
+    if use_cache:
+        # Ensure the directory exists
+        CACHE_BTC_PATH.parent.mkdir(parents=True, exist_ok=True)
 
     # Check if the file exists and was modified today
     if (
-        USE_CACHE and
+        use_cache and
         CACHE_BTC_PATH.is_file()
         and datetime.datetime.fromtimestamp(CACHE_BTC_PATH.stat().st_mtime).date()
         == datetime.datetime.today().date()
@@ -44,8 +48,8 @@ def btc_historical_daily() -> pd.DataFrame:
     return df
 
 
-def btc_pi(time_grouping: TypeTime, periods_back: int | None = None):
-    df = btc_historical_daily()[["close"]].rename({"close": "price"}, axis=1)
+def btc_pi(time_grouping: TypeTime, periods_back: int | None = None, **kwargs):
+    df = btc_historical_daily(**kwargs)[["close"]].rename({"close": "price"}, axis=1)
 
     df["sma111"] = df["price"].rolling(window=111).mean()
     df["sma350x2"] = df["price"].rolling(window=350).mean() * 2
@@ -84,8 +88,8 @@ def btc_pi(time_grouping: TypeTime, periods_back: int | None = None):
     return df
 
 
-def btc_hist(time_grouping: TypeTime | None = None, periods_back: int | None = None):
-    df = btc_historical_daily()
+def btc_hist(time_grouping: TypeTime | None = None, periods_back: int | None = None, **kwargs):
+    df = btc_historical_daily(**kwargs)
     df = df.rename(
         {"open": "Open", "close": "Close", "high": "High", "low": "Low"}, axis=1
     )
@@ -107,8 +111,9 @@ def btc_power_law(
     time_grouping: TypeTime | None = None,
     periods_back: int | None = None,
     years_to_predict: int = 20,
+    **kwargs
 ) -> pd.DataFrame:
-    df = btc_hist(time_grouping=time_grouping)[["Close"]].rename(
+    df = btc_hist(time_grouping=time_grouping, **kwargs)[["Close"]].rename(
         {"Close": "price"}, axis=1
     )