lumibot 4.2.0__py3-none-any.whl → 4.2.2__py3-none-any.whl

lumibot/backtesting/thetadata_backtesting_pandas.py

@@ -184,7 +184,7 @@ class ThetaDataBacktestingPandas(PandasData):
         asset: Optional[Asset] = None,  # DEBUG-LOG: Added for logging
     ) -> Optional[pd.DataFrame]:
         # DEBUG-LOG: Method entry with full parameter context
-        logger.info(
+        logger.debug(
             "[THETA][DEBUG][PANDAS][FINALIZE][ENTRY] asset=%s current_dt=%s requested_length=%s timeshift=%s input_shape=%s input_columns=%s input_index_type=%s input_has_tz=%s input_index_sample=%s",
             getattr(asset, 'symbol', asset) if asset else 'UNKNOWN',
             current_dt.isoformat() if hasattr(current_dt, 'isoformat') else current_dt,
@@ -199,7 +199,7 @@ class ThetaDataBacktestingPandas(PandasData):
 
         if pandas_df is None or pandas_df.empty:
             # DEBUG-LOG: Early return for empty input
-            logger.info(
+            logger.debug(
                 "[THETA][DEBUG][PANDAS][FINALIZE][EMPTY_INPUT] asset=%s returning_none_or_empty=True",
                 getattr(asset, 'symbol', asset) if asset else 'UNKNOWN'
             )
@@ -212,7 +212,7 @@ class ThetaDataBacktestingPandas(PandasData):
         frame.index = pd.to_datetime(frame.index)
 
         # DEBUG-LOG: Timezone state before localization
-        logger.info(
+        logger.debug(
             "[THETA][DEBUG][PANDAS][FINALIZE][TZ_CHECK] asset=%s frame_index_tz=%s target_tz=%s needs_localization=%s frame_shape=%s",
             getattr(asset, 'symbol', asset) if asset else 'UNKNOWN',
             frame.index.tz,
@@ -227,7 +227,7 @@ class ThetaDataBacktestingPandas(PandasData):
         normalized_for_cutoff = localized_index.normalize()
 
         # DEBUG-LOG: After localization
-        logger.info(
+        logger.debug(
             "[THETA][DEBUG][PANDAS][FINALIZE][LOCALIZED] asset=%s localized_index_tz=%s localized_sample=%s",
             getattr(asset, 'symbol', asset) if asset else 'UNKNOWN',
             localized_index.tz,
@@ -238,7 +238,7 @@ class ThetaDataBacktestingPandas(PandasData):
         cutoff_mask = normalized_for_cutoff <= cutoff
 
         # DEBUG-LOG: Cutoff filtering state
-        logger.info(
+        logger.debug(
             "[THETA][DEBUG][PANDAS][FINALIZE][CUTOFF] asset=%s cutoff=%s cutoff_mask_true=%s cutoff_mask_false=%s",
             getattr(asset, 'symbol', asset) if asset else 'UNKNOWN',
             cutoff,
@@ -249,7 +249,7 @@ class ThetaDataBacktestingPandas(PandasData):
         if timeshift and not isinstance(timeshift, int):
             cutoff_mask &= normalized_for_cutoff <= (cutoff - timeshift)
             # DEBUG-LOG: After timeshift adjustment
-            logger.info(
+            logger.debug(
                 "[THETA][DEBUG][PANDAS][FINALIZE][TIMESHIFT_ADJUSTED] asset=%s timeshift=%s new_cutoff=%s cutoff_mask_true=%s",
                 getattr(asset, 'symbol', asset) if asset else 'UNKNOWN',
                 timeshift,
@@ -262,7 +262,7 @@ class ThetaDataBacktestingPandas(PandasData):
         normalized_for_cutoff = normalized_for_cutoff[cutoff_mask]
 
         # DEBUG-LOG: After cutoff filtering
-        logger.info(
+        logger.debug(
             "[THETA][DEBUG][PANDAS][FINALIZE][AFTER_CUTOFF] asset=%s shape=%s index_range=%s",
             getattr(asset, 'symbol', asset) if asset else 'UNKNOWN',
             frame.shape,
@@ -280,7 +280,7 @@ class ThetaDataBacktestingPandas(PandasData):
         raw_frame = frame.copy()
 
         # DEBUG-LOG: After normalization
-        logger.info(
+        logger.debug(
             "[THETA][DEBUG][PANDAS][FINALIZE][NORMALIZED_INDEX] asset=%s shape=%s index_sample=%s",
             getattr(asset, 'symbol', asset) if asset else 'UNKNOWN',
             frame.shape,
@@ -291,7 +291,7 @@ class ThetaDataBacktestingPandas(PandasData):
         target_index = pd.date_range(end=expected_last_dt, periods=requested_length, freq="D", tz=self.tzinfo)
 
         # DEBUG-LOG: Target index details
-        logger.info(
+        logger.debug(
             "[THETA][DEBUG][PANDAS][FINALIZE][TARGET_INDEX] asset=%s target_length=%s target_range=%s",
             getattr(asset, 'symbol', asset) if asset else 'UNKNOWN',
             len(target_index),
@@ -304,7 +304,7 @@ class ThetaDataBacktestingPandas(PandasData):
         frame = frame.reindex(target_index)
 
         # DEBUG-LOG: After reindex
-        logger.info(
+        logger.debug(
             "[THETA][DEBUG][PANDAS][FINALIZE][AFTER_REINDEX] asset=%s shape=%s columns=%s",
             getattr(asset, 'symbol', asset) if asset else 'UNKNOWN',
             frame.shape,
@@ -318,7 +318,7 @@ class ThetaDataBacktestingPandas(PandasData):
             placeholder_mask = frame.isna().all(axis=1)
 
         # DEBUG-LOG: Placeholder mask computation
-        logger.info(
+        logger.debug(
             "[THETA][DEBUG][PANDAS][FINALIZE][PLACEHOLDER_MASK] asset=%s placeholder_true=%s placeholder_false=%s value_columns=%s",
             getattr(asset, 'symbol', asset) if asset else 'UNKNOWN',
             int(placeholder_mask.sum()) if hasattr(placeholder_mask, 'sum') else 'N/A',
@@ -359,7 +359,7 @@ class ThetaDataBacktestingPandas(PandasData):
         # DEBUG-LOG: Final missing flag state
         try:
             missing_count = int(frame["missing"].sum())
-            logger.info(
+            logger.debug(
                 "[THETA][DEBUG][PANDAS][FINALIZE][MISSING_FINAL] asset=%s missing_true=%s missing_false=%s total_rows=%s",
                 getattr(asset, 'symbol', asset) if asset else 'UNKNOWN',
                 missing_count,
@@ -367,14 +367,14 @@ class ThetaDataBacktestingPandas(PandasData):
                 len(frame)
             )
         except Exception as e:
-            logger.info(
+            logger.debug(
                 "[THETA][DEBUG][PANDAS][FINALIZE][MISSING_FINAL] asset=%s error=%s",
                 getattr(asset, 'symbol', asset) if asset else 'UNKNOWN',
                 str(e)
             )
 
         # DEBUG-LOG: Return value
-        logger.info(
+        logger.debug(
             "[THETA][DEBUG][PANDAS][FINALIZE][RETURN] asset=%s shape=%s columns=%s index_range=%s",
             getattr(asset, 'symbol', asset) if asset else 'UNKNOWN',
             frame.shape,
@@ -451,7 +451,7 @@ class ThetaDataBacktestingPandas(PandasData):
             existing_end = existing_meta.get("end")
 
             # DEBUG-LOG: Cache validation entry
-            logger.info(
+            logger.debug(
                 "[DEBUG][BACKTEST][THETA][DEBUG][PANDAS][CACHE_VALIDATION][ENTRY] asset=%s timestep=%s | "
                 "REQUESTED: start=%s start_threshold=%s end_requirement=%s length=%d | "
                 "EXISTING: start=%s end=%s rows=%d",
@@ -472,7 +472,7 @@ class ThetaDataBacktestingPandas(PandasData):
             )
 
             # DEBUG-LOG: Start validation result
-            logger.info(
+            logger.debug(
                 "[DEBUG][BACKTEST][THETA][DEBUG][PANDAS][START_VALIDATION] asset=%s | "
                 "start_ok=%s | "
                 "existing_start=%s start_threshold=%s | "
@@ -489,7 +489,7 @@ class ThetaDataBacktestingPandas(PandasData):
             end_ok = True
 
             # DEBUG-LOG: End validation entry
-            logger.info(
+            logger.debug(
                 "[DEBUG][BACKTEST][THETA][DEBUG][PANDAS][END_VALIDATION][ENTRY] asset=%s | "
                 "end_requirement=%s existing_end=%s tail_placeholder=%s",
                 asset_separated.symbol if hasattr(asset_separated, 'symbol') else str(asset_separated),
@@ -501,7 +501,7 @@ class ThetaDataBacktestingPandas(PandasData):
             if end_requirement is not None:
                 if existing_end is None:
                     end_ok = False
-                    logger.info(
+                    logger.debug(
                         "[DEBUG][BACKTEST][THETA][DEBUG][PANDAS][END_VALIDATION][RESULT] asset=%s | "
                         "end_ok=FALSE | reason=existing_end_is_None",
                         asset_separated.symbol if hasattr(asset_separated, 'symbol') else str(asset_separated)
@@ -520,7 +520,7 @@ class ThetaDataBacktestingPandas(PandasData):
 
                     if existing_end_cmp > end_requirement_cmp:
                         end_ok = True
-                        logger.info(
+                        logger.debug(
                             "[DEBUG][BACKTEST][THETA][DEBUG][PANDAS][END_VALIDATION][RESULT] asset=%s | "
                             "end_ok=TRUE | reason=existing_end_exceeds_requirement | "
                             "existing_end=%s end_requirement=%s ts_unit=%s",
@@ -535,7 +535,7 @@ class ThetaDataBacktestingPandas(PandasData):
                         placeholder_empty_fetch = tail_placeholder and existing_meta.get("empty_fetch")
                         end_ok = (not tail_placeholder) or placeholder_on_weekend or placeholder_empty_fetch
 
-                        logger.info(
+                        logger.debug(
                             "[DEBUG][BACKTEST][THETA][DEBUG][PANDAS][END_VALIDATION][EXACT_MATCH] asset=%s | "
                             "existing_end == end_requirement | "
                             "weekday=%s placeholder_on_weekend=%s placeholder_empty_fetch=%s | "
@@ -549,7 +549,7 @@ class ThetaDataBacktestingPandas(PandasData):
                         )
                     else:
                         end_ok = False
-                        logger.info(
+                        logger.debug(
                             "[DEBUG][BACKTEST][THETA][DEBUG][PANDAS][END_VALIDATION][RESULT] asset=%s | "
                             "end_ok=FALSE | reason=existing_end_less_than_requirement | "
                             "existing_end=%s end_requirement=%s ts_unit=%s",
@@ -566,7 +566,7 @@ class ThetaDataBacktestingPandas(PandasData):
             )
 
             # DEBUG-LOG: Final cache decision
-            logger.info(
+            logger.debug(
                 "[DEBUG][BACKTEST][THETA][DEBUG][PANDAS][CACHE_DECISION] asset=%s | "
                 "cache_covers=%s | "
                 "start_ok=%s rows_ok=%s (existing=%d >= requested=%d) end_ok=%s",
@@ -586,7 +586,7 @@ class ThetaDataBacktestingPandas(PandasData):
                     and expiration_dt == end_requirement
                     and not existing_meta.get("expiration_notice")
                 ):
-                    logger.info(
+                    logger.debug(
                         "[THETA][DEBUG][THETADATA-PANDAS] Reusing cached data for %s/%s through option expiry %s.",
                         asset_separated,
                         quote_asset,
@@ -702,7 +702,7 @@ class ThetaDataBacktestingPandas(PandasData):
                 and expiration_dt == end_requirement
             )
             if expired_reason:
-                logger.info(
+                logger.debug(
                     "[THETA][DEBUG][THETADATA-PANDAS] No new OHLC rows for %s/%s (%s); option expired on %s. Keeping cached data.",
                     asset_separated,
                     quote_asset,
@@ -851,9 +851,8 @@ class ThetaDataBacktestingPandas(PandasData):
         bars = self._parse_source_symbol_bars(response, asset, quote=quote)
         final_df = getattr(bars, "df", None)
         final_rows = len(final_df) if final_df is not None else 0
-        message = (
-            "[THETA][DEBUG][FETCH][THETA][DEBUG][PANDAS][FINAL] asset=%s quote=%s length=%s timestep=%s timeshift=%s current_dt=%s rows=%s"
-        ) % (
+        logger.debug(
+            "[THETA][DEBUG][FETCH][THETA][DEBUG][PANDAS][FINAL] asset=%s quote=%s length=%s timestep=%s timeshift=%s current_dt=%s rows=%s",
             getattr(asset, "symbol", asset) if not isinstance(asset, str) else asset,
             getattr(quote, "symbol", quote),
             length,
@@ -862,8 +861,6 @@ class ThetaDataBacktestingPandas(PandasData):
             current_dt,
             final_rows,
         )
-        logger.warning(message)
-        print(message)
         return bars
 
     def get_last_price(self, asset, timestep="minute", quote=None, exchange=None, **kwargs) -> Union[float, Decimal, None]:
@@ -893,7 +890,7 @@ class ThetaDataBacktestingPandas(PandasData):
                     return super().get_last_price(asset=asset, quote=quote, exchange=exchange)
                 closes = close_series.dropna()
                 if closes.empty:
-                    logger.warning(
+                    logger.debug(
                         "[THETA][DEBUG][THETADATA-PANDAS] get_last_price found no valid closes for %s/%s; returning None (likely expired).",
                         asset,
                         quote or Asset("USD", "forex"),
@@ -957,10 +954,9 @@ class ThetaDataBacktestingPandas(PandasData):
             return_polars=False,
         )
         if bars is None or getattr(bars, "df", None) is None or bars.df.empty:
-            message = (
+            logger.debug(
                 "[THETA][DEBUG][FETCH][THETA][DEBUG][PANDAS] asset=%s quote=%s length=%s timestep=%s timeshift=%s current_dt=%s "
-                "rows=0 first_ts=None last_ts=None columns=None"
-            ) % (
+                "rows=0 first_ts=None last_ts=None columns=None",
                 getattr(asset, "symbol", asset) if not isinstance(asset, str) else asset,
                 getattr(quote, "symbol", quote),
                 length,
@@ -968,8 +964,6 @@ class ThetaDataBacktestingPandas(PandasData):
                 timeshift,
                 current_dt,
             )
-            logger.warning(message)
-            print(message)
             return bars
 
         df = bars.df
@@ -981,10 +975,10 @@ class ThetaDataBacktestingPandas(PandasData):
         else:
             first_ts = df.index[0]
             last_ts = df.index[-1]
-        message = (
+
+        logger.debug(
             "[THETA][DEBUG][FETCH][THETA][DEBUG][PANDAS] asset=%s quote=%s length=%s timestep=%s timeshift=%s current_dt=%s rows=%s "
-            "first_ts=%s last_ts=%s columns=%s"
-        ) % (
+            "first_ts=%s last_ts=%s columns=%s",
             getattr(asset, "symbol", asset) if not isinstance(asset, str) else asset,
             getattr(quote, "symbol", quote),
             length,
@@ -996,8 +990,6 @@ class ThetaDataBacktestingPandas(PandasData):
             last_ts,
             columns,
         )
-        logger.warning(message)
-        print(message)
         return bars
 
     def get_quote(self, asset, timestep="minute", quote=None, exchange=None, **kwargs):
@@ -1026,7 +1018,7 @@ class ThetaDataBacktestingPandas(PandasData):
 
         # [INSTRUMENTATION] Log full asset details for options
         if hasattr(asset, 'asset_type') and asset.asset_type == Asset.AssetType.OPTION:
-            logger.info(
+            logger.debug(
                 "[THETA][DEBUG][QUOTE][THETA][DEBUG][PANDAS][OPTION_REQUEST] symbol=%s expiration=%s strike=%s right=%s current_dt=%s timestep=%s",
                 asset.symbol,
                 asset.expiration,
@@ -1036,7 +1028,7 @@ class ThetaDataBacktestingPandas(PandasData):
                 timestep
             )
         else:
-            logger.info(
+            logger.debug(
                 "[THETA][DEBUG][QUOTE][THETA][DEBUG][PANDAS][REQUEST] asset=%s current_dt=%s timestep=%s",
                 getattr(asset, "symbol", asset) if not isinstance(asset, str) else asset,
                 dt.isoformat() if hasattr(dt, 'isoformat') else dt,
@@ -1066,7 +1058,7 @@ class ThetaDataBacktestingPandas(PandasData):
                 if isinstance(df.index, pd.DatetimeIndex) and df.index.tz is not None:
                     tz_info = str(df.index.tz)
 
-                logger.info(
+                logger.debug(
                     "[THETA][DEBUG][QUOTE][THETA][DEBUG][PANDAS][DATAFRAME_STATE] asset=%s | total_rows=%d | timestep=%s | index_type=%s | timezone=%s",
                     getattr(asset, "symbol", asset),
                     len(df),
@@ -1079,7 +1071,7 @@ class ThetaDataBacktestingPandas(PandasData):
                 if isinstance(df.index, pd.DatetimeIndex):
                     first_dt_str = df.index[0].isoformat() if hasattr(df.index[0], 'isoformat') else str(df.index[0])
                     last_dt_str = df.index[-1].isoformat() if hasattr(df.index[-1], 'isoformat') else str(df.index[-1])
-                    logger.info(
+                    logger.debug(
                         "[THETA][DEBUG][QUOTE][THETA][DEBUG][PANDAS][DATETIME_RANGE] asset=%s | first_dt=%s | last_dt=%s | tz=%s",
                         getattr(asset, "symbol", asset),
                         first_dt_str,
@@ -1089,12 +1081,12 @@ class ThetaDataBacktestingPandas(PandasData):
 
                     # CRITICAL: Show tail with explicit datetime index to catch time-travel bug
                     if debug_enabled and len(available_cols) > 0:
-                        logger.info(
+                        logger.debug(
                             "[THETA][DEBUG][QUOTE][THETA][DEBUG][PANDAS][DATAFRAME_HEAD] asset=%s | first_5_rows (with datetime index):\n%s",
                             getattr(asset, "symbol", asset),
                             head_df[available_cols].to_string()
                         )
-                        logger.info(
+                        logger.debug(
                             "[THETA][DEBUG][QUOTE][THETA][DEBUG][PANDAS][DATAFRAME_TAIL] asset=%s | last_5_rows (with datetime index):\n%s",
                             getattr(asset, "symbol", asset),
                             tail_df[available_cols].to_string()
@@ -1102,18 +1094,18 @@ class ThetaDataBacktestingPandas(PandasData):
 
                         # Show tail datetime values explicitly
                         tail_datetimes = [dt.isoformat() if hasattr(dt, 'isoformat') else str(dt) for dt in tail_df.index]
-                        logger.info(
+                        logger.debug(
                             "[THETA][DEBUG][QUOTE][THETA][DEBUG][PANDAS][TAIL_DATETIMES] asset=%s | tail_index=%s",
                             getattr(asset, "symbol", asset),
                             tail_datetimes
                         )
             else:
-                logger.info(
+                logger.debug(
                     "[THETA][DEBUG][QUOTE][THETA][DEBUG][PANDAS][DATAFRAME_STATE] asset=%s | EMPTY_DATAFRAME",
                     getattr(asset, "symbol", asset)
                 )
         else:
-            logger.info(
+            logger.debug(
                 "[THETA][DEBUG][QUOTE][THETA][DEBUG][PANDAS][DATAFRAME_STATE] asset=%s | NO_DATA_FOUND_IN_STORE",
                 getattr(asset, "symbol", asset)
             )
@@ -1121,9 +1113,8 @@ class ThetaDataBacktestingPandas(PandasData):
         quote_obj = super().get_quote(asset=asset, quote=quote, exchange=exchange)
 
         # [INSTRUMENTATION] Final quote result with all details
-        message = (
-            "[THETA][DEBUG][QUOTE][THETA][DEBUG][PANDAS][RESULT] asset=%s quote=%s current_dt=%s bid=%s ask=%s mid=%s last=%s source=%s"
-        ) % (
+        logger.debug(
+            "[THETA][DEBUG][QUOTE][THETA][DEBUG][PANDAS][RESULT] asset=%s quote=%s current_dt=%s bid=%s ask=%s mid=%s last=%s source=%s",
             getattr(asset, "symbol", asset) if not isinstance(asset, str) else asset,
             getattr(quote, "symbol", quote),
             dt,
@@ -1133,8 +1124,6 @@ class ThetaDataBacktestingPandas(PandasData):
             getattr(quote_obj, "last_price", None) if quote_obj else None,
             getattr(quote_obj, "source", None) if quote_obj else None,
         )
-        logger.warning(message)
-        print(message)
         return quote_obj
 
     def get_chains(self, asset):

lumibot/tools/thetadata_helper.py

@@ -26,6 +26,17 @@ CONNECTION_MAX_RETRIES = 60
 BOOT_GRACE_PERIOD = 5.0
 MAX_RESTART_ATTEMPTS = 3
 
+
+def _resolve_asset_folder(asset_obj: Asset) -> str:
+    asset_type = getattr(asset_obj, "asset_type", None) or "stock"
+    asset_key = str(asset_type).strip().lower()
+    return asset_key
+
+
+def _normalize_folder_component(value: str, fallback: str) -> str:
+    normalized = str(value or "").strip().lower().replace(" ", "_")
+    return normalized or fallback
+
 # Global process tracking for ThetaTerminal
 THETA_DATA_PROCESS = None
 THETA_DATA_PID = None
@@ -98,7 +109,7 @@ def append_missing_markers(
         CONNECTION_DIAGNOSTICS["placeholder_writes"] = CONNECTION_DIAGNOSTICS.get("placeholder_writes", 0) + len(rows)
 
         # DEBUG-LOG: Placeholder injection
-        logger.info(
+        logger.debug(
             "[THETA][DEBUG][PLACEHOLDER][INJECT] count=%d dates=%s",
             len(rows),
             ", ".join(sorted({d.isoformat() for d in missing_dates}))
@@ -114,7 +125,7 @@ def append_missing_markers(
         else:
             df_all = pd.concat([df_all, placeholder_df]).sort_index()
         df_all = df_all[~df_all.index.duplicated(keep="last")]
-        logger.info(
+        logger.debug(
             "[THETA][DEBUG][THETADATA-CACHE] recorded %d placeholder day(s): %s",
             len(rows),
             ", ".join(sorted({d.isoformat() for d in missing_dates})),
@@ -140,7 +151,7 @@ def remove_missing_markers(
     if mask.any():
         removed_dates = sorted({ts.date().isoformat() for ts in df_all.index[mask]})
         df_all = df_all.loc[~mask]
-        logger.info(
+        logger.debug(
             "[THETA][DEBUG][THETADATA-CACHE] cleared %d placeholder row(s) for dates: %s",
             mask.sum(),
             ", ".join(removed_dates),
@@ -274,7 +285,7 @@ def get_price_data(
         try:
             fetched_remote = cache_manager.ensure_local_file(cache_file, payload=remote_payload)
             if fetched_remote:
-                logger.info(
+                logger.debug(
                     "[THETA][DEBUG][CACHE][REMOTE_DOWNLOAD] asset=%s timespan=%s datastyle=%s cache_file=%s",
                     asset,
                     timespan,
@@ -282,7 +293,7 @@ def get_price_data(
                     cache_file,
                 )
         except Exception as exc:
-            logger.exception(
+            logger.debug(
                 "[THETA][DEBUG][CACHE][REMOTE_DOWNLOAD_ERROR] asset=%s cache_file=%s error=%s",
                 asset,
                 cache_file,
@@ -290,7 +301,7 @@ def get_price_data(
             )
 
     # DEBUG-LOG: Cache file check
-    logger.info(
+    logger.debug(
         "[THETA][DEBUG][CACHE][CHECK] asset=%s timespan=%s datastyle=%s cache_file=%s exists=%s",
         asset,
         timespan,
@@ -311,7 +322,7 @@ def get_price_data(
         placeholder_rows = int(df_all["missing"].sum())
 
     # DEBUG-LOG: Cache load result
-    logger.info(
+    logger.debug(
         "[THETA][DEBUG][CACHE][LOADED] asset=%s cached_rows=%d placeholder_rows=%d real_rows=%d",
         asset,
         cached_rows,
@@ -329,7 +340,7 @@ def get_price_data(
     )
 
     # Check if we need to get more data
-    logger.info(
+    logger.debug(
         "[THETA][DEBUG][CACHE][DECISION_START] asset=%s | "
         "calling get_missing_dates(start=%s, end=%s)",
         asset.symbol if hasattr(asset, 'symbol') else str(asset),
@@ -339,7 +350,7 @@ def get_price_data(
 
     missing_dates = get_missing_dates(df_all, asset, start, end)
 
-    logger.info(
+    logger.debug(
         "[THETA][DEBUG][CACHE][DECISION_RESULT] asset=%s | "
         "missing_dates=%d | "
         "decision=%s",
@@ -363,7 +374,7 @@ def get_price_data(
         if df_all is not None and not df_all.empty:
             logger.info("ThetaData cache HIT for %s %s %s (%d rows).", asset, timespan, datastyle, len(df_all))
             # DEBUG-LOG: Cache hit
-            logger.info(
+            logger.debug(
                 "[THETA][DEBUG][CACHE][HIT] asset=%s timespan=%s datastyle=%s rows=%d start=%s end=%s",
                 asset,
                 timespan,
@@ -390,7 +401,7 @@ def get_price_data(
                 # DEBUG-LOG: Entry to intraday filter
                 rows_before_any_filter = len(df_all)
                 max_ts_before_any_filter = df_all.index.max() if len(df_all) > 0 else None
-                logger.info(
+                logger.debug(
                     "[THETA][DEBUG][FILTER][INTRADAY_ENTRY] asset=%s | "
                     "rows_before=%d max_ts_before=%s | "
                     "start_param=%s end_param=%s dt_param=%s dt_type=%s",
@@ -406,13 +417,13 @@ def get_price_data(
                 # Convert date to datetime if needed
                 if isinstance(start, datetime_module.date) and not isinstance(start, datetime_module.datetime):
                     start = datetime_module.datetime.combine(start, datetime_module.time.min)
-                    logger.info(
+                    logger.debug(
                         "[THETA][DEBUG][FILTER][DATE_CONVERSION] converted start from date to datetime: %s",
                         start.isoformat()
                     )
                 if isinstance(end, datetime_module.date) and not isinstance(end, datetime_module.datetime):
                     end = datetime_module.datetime.combine(end, datetime_module.time.max)
-                    logger.info(
+                    logger.debug(
                         "[THETA][DEBUG][FILTER][DATE_CONVERSION] converted end from date to datetime: %s",
                         end.isoformat()
                     )
@@ -421,20 +432,20 @@ def get_price_data(
                 if isinstance(end, datetime_module.datetime) and end.time() == datetime_module.time.min:
                     # Convert end-of-period midnight to end-of-day
                     end = datetime_module.datetime.combine(end.date(), datetime_module.time.max)
-                    logger.info(
+                    logger.debug(
                         "[THETA][DEBUG][FILTER][MIDNIGHT_FIX] converted end from midnight to end-of-day: %s",
                         end.isoformat()
                     )
 
                 if start.tzinfo is None:
                     start = LUMIBOT_DEFAULT_PYTZ.localize(start).astimezone(pytz.UTC)
-                    logger.info(
+                    logger.debug(
                         "[THETA][DEBUG][FILTER][TZ_LOCALIZE] localized start to UTC: %s",
                         start.isoformat()
                     )
                 if end.tzinfo is None:
                     end = LUMIBOT_DEFAULT_PYTZ.localize(end).astimezone(pytz.UTC)
-                    logger.info(
+                    logger.debug(
                         "[THETA][DEBUG][FILTER][TZ_LOCALIZE] localized end to UTC: %s",
                         end.isoformat()
                     )
@@ -445,7 +456,7 @@ def get_price_data(
                 #
                 # NEW APPROACH: Always return full [start, end] range from cache
                 # Let Data/DataPolars.get_bars() handle look-ahead bias protection
-                logger.info(
+                logger.debug(
                     "[THETA][DEBUG][FILTER][NO_DT_FILTER] asset=%s | "
                     "using end=%s for upper bound (dt parameter ignored for cache retrieval)",
                     asset.symbol if hasattr(asset, 'symbol') else str(asset),
@@ -455,7 +466,7 @@ def get_price_data(
 
         # DEBUG-LOG: After date range filtering, before missing removal
         if df_all is not None and not df_all.empty:
-            logger.info(
+            logger.debug(
                 "[THETA][DEBUG][FILTER][AFTER] asset=%s rows=%d first_ts=%s last_ts=%s dt_filter=%s",
                 asset,
                 len(df_all),
@@ -470,7 +481,7 @@ def get_price_data(
 
         # DEBUG-LOG: Before pandas return
         if df_all is not None and not df_all.empty:
-            logger.info(
+            logger.debug(
                 "[THETA][DEBUG][RETURN][PANDAS] asset=%s rows=%d first_ts=%s last_ts=%s",
                 asset,
                 len(df_all),
@@ -482,7 +493,7 @@ def get_price_data(
     logger.info("ThetaData cache MISS for %s %s %s; fetching %d interval(s) from ThetaTerminal.", asset, timespan, datastyle, len(missing_dates))
 
     # DEBUG-LOG: Cache miss
-    logger.info(
+    logger.debug(
         "[THETA][DEBUG][CACHE][MISS] asset=%s timespan=%s datastyle=%s missing_intervals=%d first=%s last=%s",
         asset,
         timespan,
@@ -542,7 +553,7 @@ def get_price_data(
                 and all(day > asset.expiration for day in requested_dates)
             )
             if expired_range:
-                logger.info(
+                logger.debug(
                     "[THETA][DEBUG][THETADATA-EOD] Option %s expired on %s; cache reuse for range %s -> %s.",
                     asset,
                     asset.expiration,
@@ -550,7 +561,7 @@ def get_price_data(
                     fetch_end,
                 )
             else:
-                logger.warning(
+                logger.debug(
                     "[THETA][DEBUG][THETADATA-EOD] No rows returned for %s between %s and %s; recording placeholders.",
                     asset,
                     fetch_start,
@@ -680,7 +691,7 @@ def get_price_data(
                 and chunk_end.date() >= asset.expiration
             )
             if expired_chunk:
-                logger.info(
+                logger.debug(
                     "[THETA][DEBUG][THETADATA] Option %s considered expired on %s; reusing cached data between %s and %s.",
                     asset,
                     asset.expiration,
@@ -785,7 +796,11 @@ def get_trading_dates(asset: Asset, start: datetime, end: datetime):
 def build_cache_filename(asset: Asset, timespan: str, datastyle: str = "ohlc"):
     """Helper function to create the cache filename for a given asset and timespan"""
 
-    lumibot_cache_folder = Path(LUMIBOT_CACHE_FOLDER) / CACHE_SUBFOLDER
+    provider_root = Path(LUMIBOT_CACHE_FOLDER) / CACHE_SUBFOLDER
+    asset_folder = _resolve_asset_folder(asset)
+    timespan_folder = _normalize_folder_component(timespan, "unknown")
+    datastyle_folder = _normalize_folder_component(datastyle, "default")
+    base_folder = provider_root / asset_folder / timespan_folder / datastyle_folder
 
     # If It's an option then also add the expiration date, strike price and right to the filename
     if asset.asset_type == "option":
@@ -799,7 +814,7 @@ def build_cache_filename(asset: Asset, timespan: str, datastyle: str = "ohlc"):
         uniq_str = asset.symbol
 
     cache_filename = f"{asset.asset_type}_{uniq_str}_{timespan}_{datastyle}.parquet"
-    cache_file = lumibot_cache_folder / cache_filename
+    cache_file = base_folder / cache_filename
     return cache_file
 
 
@@ -848,7 +863,7 @@ def get_missing_dates(df_all, asset, start, end):
         A list of dates that we need to get data for
     """
     # DEBUG-LOG: Entry to get_missing_dates
-    logger.info(
+    logger.debug(
         "[THETA][DEBUG][CACHE][MISSING_DATES_CHECK] asset=%s | "
         "start=%s end=%s | "
         "cache_rows=%d",
@@ -860,7 +875,7 @@ def get_missing_dates(df_all, asset, start, end):
 
     trading_dates = get_trading_dates(asset, start, end)
 
-    logger.info(
+    logger.debug(
         "[THETA][DEBUG][CACHE][TRADING_DATES] asset=%s | "
         "trading_dates_count=%d first=%s last=%s",
         asset.symbol if hasattr(asset, 'symbol') else str(asset),
@@ -870,7 +885,7 @@ def get_missing_dates(df_all, asset, start, end):
     )
 
     if df_all is None or not len(df_all):
-        logger.info(
+        logger.debug(
             "[THETA][DEBUG][CACHE][EMPTY] asset=%s | "
             "cache is EMPTY -> all %d trading days are missing",
             asset.symbol if hasattr(asset, 'symbol') else str(asset),
@@ -886,7 +901,7 @@ def get_missing_dates(df_all, asset, start, end):
     cached_first = min(dates) if len(dates) > 0 else None
     cached_last = max(dates) if len(dates) > 0 else None
 
-    logger.info(
+    logger.debug(
         "[THETA][DEBUG][CACHE][CACHED_DATES] asset=%s | "
         "cached_dates_count=%d first=%s last=%s",
         asset.symbol if hasattr(asset, 'symbol') else str(asset),
@@ -904,7 +919,7 @@ def get_missing_dates(df_all, asset, start, end):
         after_expiry_filter = len(missing_dates)
 
         if before_expiry_filter != after_expiry_filter:
-            logger.info(
+            logger.debug(
                 "[THETA][DEBUG][CACHE][OPTION_EXPIRY_FILTER] asset=%s | "
                 "filtered %d dates after expiration=%s | "
                 "missing_dates: %d -> %d",
@@ -915,7 +930,7 @@ def get_missing_dates(df_all, asset, start, end):
                 after_expiry_filter
             )
 
-    logger.info(
+    logger.debug(
         "[THETA][DEBUG][CACHE][MISSING_RESULT] asset=%s | "
         "missing_dates_count=%d | "
         "first_missing=%s last_missing=%s",
@@ -931,7 +946,7 @@ def get_missing_dates(df_all, asset, start, end):
 def load_cache(cache_file):
     """Load the data from the cache file and return a DataFrame with a DateTimeIndex"""
     # DEBUG-LOG: Start loading cache
-    logger.info(
+    logger.debug(
         "[THETA][DEBUG][CACHE][LOAD_START] cache_file=%s | "
         "exists=%s size_bytes=%d",
         cache_file.name,
@@ -940,7 +955,7 @@ def load_cache(cache_file):
     )
 
     if not cache_file.exists():
-        logger.info(
+        logger.debug(
             "[THETA][DEBUG][CACHE][LOAD_MISSING] cache_file=%s | returning=None",
             cache_file.name,
         )
@@ -949,7 +964,7 @@ def load_cache(cache_file):
     df = pd.read_parquet(cache_file, engine='pyarrow')
 
     rows_after_read = len(df)
-    logger.info(
+    logger.debug(
         "[THETA][DEBUG][CACHE][LOAD_READ] cache_file=%s | "
         "rows_read=%d columns=%s",
         cache_file.name,
@@ -969,7 +984,7 @@ def load_cache(cache_file):
     if df.index.tzinfo is None:
         # Set the timezone to UTC
         df.index = df.index.tz_localize("UTC")
-        logger.info(
+        logger.debug(
             "[THETA][DEBUG][CACHE][LOAD_TZ] cache_file=%s | "
             "localized index to UTC",
             cache_file.name
@@ -981,7 +996,7 @@ def load_cache(cache_file):
     max_ts = df.index.max() if len(df) > 0 else None
     placeholder_count = int(df["missing"].sum()) if "missing" in df.columns else 0
 
-    logger.info(
+    logger.debug(
         "[THETA][DEBUG][CACHE][LOAD_SUCCESS] cache_file=%s | "
         "total_rows=%d real_rows=%d placeholders=%d | "
         "min_ts=%s max_ts=%s",
@@ -999,7 +1014,7 @@ def load_cache(cache_file):
 def update_cache(cache_file, df_all, df_cached, missing_dates=None, remote_payload=None):
     """Update the cache file with the new data and optional placeholder markers."""
     # DEBUG-LOG: Entry to update_cache
-    logger.info(
+    logger.debug(
         "[THETA][DEBUG][CACHE][UPDATE_ENTRY] cache_file=%s | "
         "df_all_rows=%d df_cached_rows=%d missing_dates=%d",
         cache_file.name,
@@ -1010,13 +1025,13 @@ def update_cache(cache_file, df_all, df_cached, missing_dates=None, remote_paylo
 
     if df_all is None or len(df_all) == 0:
         if not missing_dates:
-            logger.info(
+            logger.debug(
                 "[THETA][DEBUG][CACHE][UPDATE_SKIP] cache_file=%s | "
                 "df_all is empty and no missing_dates, skipping cache update",
                 cache_file.name
             )
             return
-        logger.info(
+        logger.debug(
             "[THETA][DEBUG][CACHE][UPDATE_PLACEHOLDERS_ONLY] cache_file=%s | "
             "df_all is empty, writing %d placeholders",
             cache_file.name,
@@ -1026,7 +1041,7 @@ def update_cache(cache_file, df_all, df_cached, missing_dates=None, remote_paylo
     else:
         df_working = ensure_missing_column(df_all.copy())
         if missing_dates:
-            logger.info(
+            logger.debug(
                 "[THETA][DEBUG][CACHE][UPDATE_APPEND_PLACEHOLDERS] cache_file=%s | "
                 "appending %d placeholders to %d existing rows",
                 cache_file.name,
@@ -1036,7 +1051,7 @@ def update_cache(cache_file, df_all, df_cached, missing_dates=None, remote_paylo
             df_working = append_missing_markers(df_working, missing_dates)
 
     if df_working is None or len(df_working) == 0:
-        logger.info(
+        logger.debug(
             "[THETA][DEBUG][CACHE][UPDATE_SKIP_EMPTY] cache_file=%s | "
             "df_working is empty after processing, skipping write",
             cache_file.name
@@ -1048,7 +1063,7 @@ def update_cache(cache_file, df_all, df_cached, missing_dates=None, remote_paylo
         df_cached_cmp = ensure_missing_column(df_cached.copy())
 
     if df_cached_cmp is not None and df_working.equals(df_cached_cmp):
-        logger.info(
+        logger.debug(
             "[THETA][DEBUG][CACHE][UPDATE_NO_CHANGES] cache_file=%s | "
             "df_working equals df_cached (rows=%d), skipping write",
             cache_file.name,
@@ -1069,7 +1084,7 @@ def update_cache(cache_file, df_all, df_cached, missing_dates=None, remote_paylo
             return None
         return value.isoformat() if hasattr(value, "isoformat") else value
 
-    logger.info(
+    logger.debug(
         "[THETA][DEBUG][CACHE][UPDATE_WRITE] cache_file=%s | "
         "total_rows=%d real_rows=%d placeholders=%d | "
         "min_ts=%s max_ts=%s",
@@ -1083,7 +1098,7 @@ def update_cache(cache_file, df_all, df_cached, missing_dates=None, remote_paylo
 
     df_to_save.to_parquet(cache_file, engine="pyarrow", compression="snappy")
 
-    logger.info(
+    logger.debug(
         "[THETA][DEBUG][CACHE][UPDATE_SUCCESS] cache_file=%s written successfully",
         cache_file.name
     )
@@ -1093,7 +1108,7 @@ def update_cache(cache_file, df_all, df_cached, missing_dates=None, remote_paylo
         try:
             cache_manager.on_local_update(cache_file, payload=remote_payload)
         except Exception as exc:
-            logger.exception(
+            logger.debug(
                 "[THETA][DEBUG][CACHE][REMOTE_UPLOAD_ERROR] cache_file=%s error=%s",
                 cache_file,
                 exc,
@@ -1469,7 +1484,7 @@ def get_request(url: str, headers: dict, querystring: dict, username: str, passw
                 CONNECTION_DIAGNOSTICS["network_requests"] += 1
 
                 # DEBUG-LOG: API request
-                logger.info(
+                logger.debug(
                     "[THETA][DEBUG][API][REQUEST] url=%s params=%s",
                     request_url if next_page_url else url,
                     request_params if request_params else querystring
@@ -1480,7 +1495,7 @@ def get_request(url: str, headers: dict, querystring: dict, username: str, passw
                 if response.status_code == 472:
                     logger.warning(f"No data available for request: {response.text[:200]}")
                     # DEBUG-LOG: API response - no data
-                    logger.info(
+                    logger.debug(
                         "[THETA][DEBUG][API][RESPONSE] status=472 result=NO_DATA"
                     )
                     return None
@@ -1488,7 +1503,7 @@ def get_request(url: str, headers: dict, querystring: dict, username: str, passw
                 elif response.status_code != 200:
                     logger.warning(f"Non-200 status code {response.status_code}: {response.text[:200]}")
                     # DEBUG-LOG: API response - error
-                    logger.info(
+                    logger.debug(
                         "[THETA][DEBUG][API][RESPONSE] status=%d result=ERROR",
                         response.status_code
                     )
@@ -1498,7 +1513,7 @@ def get_request(url: str, headers: dict, querystring: dict, username: str, passw
 
                     # DEBUG-LOG: API response - success
                     response_rows = len(json_resp.get("response", [])) if isinstance(json_resp.get("response"), list) else 0
-                    logger.info(
+                    logger.debug(
                         "[THETA][DEBUG][API][RESPONSE] status=200 rows=%d has_next_page=%s",
                         response_rows,
                         bool(json_resp.get("header", {}).get("next_page"))
@@ -1524,7 +1539,7 @@ def get_request(url: str, headers: dict, querystring: dict, username: str, passw
                 logger.warning(f"Exception during request (attempt {counter + 1}): {e}")
                 check_connection(username=username, password=password, wait_for_connection=True)
                 if counter == 0:
-                    logger.info("[THETA][DEBUG][API][WAIT] Allowing ThetaTerminal to initialize for 5s before retry.")
+                    logger.debug("[THETA][DEBUG][API][WAIT] Allowing ThetaTerminal to initialize for 5s before retry.")
                     time.sleep(5)
 
             counter += 1
@@ -1609,7 +1624,7 @@ def get_historical_eod_data(asset: Asset, start_dt: datetime, end_dt: datetime,
     headers = {"Accept": "application/json"}
 
     # DEBUG-LOG: EOD data request
-    logger.info(
+    logger.debug(
         "[THETA][DEBUG][EOD][REQUEST] asset=%s start=%s end=%s datastyle=%s",
         asset,
         start_date,
@@ -1622,7 +1637,7 @@ def get_historical_eod_data(asset: Asset, start_dt: datetime, end_dt: datetime,
                             username=username, password=password)
     if json_resp is None:
         # DEBUG-LOG: EOD data response - no data
-        logger.info(
+        logger.debug(
             "[THETA][DEBUG][EOD][RESPONSE] asset=%s result=NO_DATA",
             asset
         )
@@ -1630,7 +1645,7 @@ def get_historical_eod_data(asset: Asset, start_dt: datetime, end_dt: datetime,
 
     # DEBUG-LOG: EOD data response - success
     response_rows = len(json_resp.get("response", [])) if isinstance(json_resp.get("response"), list) else 0
-    logger.info(
+    logger.debug(
         "[THETA][DEBUG][EOD][RESPONSE] asset=%s rows=%d",
         asset,
         response_rows
@@ -1786,7 +1801,7 @@ def get_historical_data(asset: Asset, start_dt: datetime, end_dt: datetime, ivl:
     headers = {"Accept": "application/json"}
 
     # DEBUG-LOG: Intraday data request
-    logger.info(
+    logger.debug(
         "[THETA][DEBUG][INTRADAY][REQUEST] asset=%s start=%s end=%s ivl=%d datastyle=%s include_after_hours=%s",
         asset,
         start_date,
@@ -1802,7 +1817,7 @@ def get_historical_data(asset: Asset, start_dt: datetime, end_dt: datetime, ivl:
                             username=username, password=password)
     if json_resp is None:
         # DEBUG-LOG: Intraday data response - no data
-        logger.info(
+        logger.debug(
             "[THETA][DEBUG][INTRADAY][RESPONSE] asset=%s result=NO_DATA",
             asset
         )
@@ -1810,7 +1825,7 @@ def get_historical_data(asset: Asset, start_dt: datetime, end_dt: datetime, ivl:
 
     # DEBUG-LOG: Intraday data response - success
     response_rows = len(json_resp.get("response", [])) if isinstance(json_resp.get("response"), list) else 0
-    logger.info(
+    logger.debug(
         "[THETA][DEBUG][INTRADAY][RESPONSE] asset=%s rows=%d",
         asset,
         response_rows
@@ -1969,7 +1984,7 @@ def get_chains_cached(
     Retrieve option chain with caching (MATCHES POLYGON PATTERN).
 
     This function follows the EXACT same caching strategy as Polygon:
-    1. Check cache: LUMIBOT_CACHE_FOLDER/thetadata/option_chains/{symbol}_{date}.parquet
+    1. Check cache: LUMIBOT_CACHE_FOLDER/thetadata/<asset-type>/option_chains/{symbol}_{date}.parquet
     2. Reuse files within RECENT_FILE_TOLERANCE_DAYS (default 7 days)
     3. If not found, fetch from ThetaData and save to cache
     4. Use pyarrow engine with snappy compression
@@ -2006,7 +2021,7 @@ def get_chains_cached(
         return None
 
     # 2) Build cache folder path
-    chain_folder = Path(LUMIBOT_CACHE_FOLDER) / "thetadata" / "option_chains"
+    chain_folder = Path(LUMIBOT_CACHE_FOLDER) / "thetadata" / _resolve_asset_folder(asset) / "option_chains"
     chain_folder.mkdir(parents=True, exist_ok=True)
 
     # 3) Check for recent cached file (within RECENT_FILE_TOLERANCE_DAYS)

{lumibot-4.2.0.dist-info → lumibot-4.2.2.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: lumibot
-Version: 4.2.0
+Version: 4.2.2
 Summary: Backtesting and Trading Library, Made by Lumiwealth
 Home-page: https://github.com/Lumiwealth/lumibot
 Author: Robert Grzesik

{lumibot-4.2.0.dist-info → lumibot-4.2.2.dist-info}/RECORD

@@ -14,7 +14,7 @@ lumibot/backtesting/interactive_brokers_rest_backtesting.py,sha256=5HJ_sPX0uOUg-
 lumibot/backtesting/pandas_backtesting.py,sha256=m-NvT4o-wFQjaZft6TXULzeZBrskO_7Z-jfy9AIkyAY,388
 lumibot/backtesting/polygon_backtesting.py,sha256=u9kif_2_7k0P4-KDvbHhaMfSoBVejUUX7fh9H3PCVE0,12350
 lumibot/backtesting/thetadata_backtesting.py,sha256=Xcz5f-4zTkKgWWcktNzItH2vrr8CysIMQWKKqLwugbA,345
-lumibot/backtesting/thetadata_backtesting_pandas.py,sha256=rglyUgZc0bNt9IWuPvmJJNXyU80JvQvRiUXWCzR1oDA,51994
+lumibot/backtesting/thetadata_backtesting_pandas.py,sha256=14XMsbQCa3uE_iS2nvTlGUXkn9kvI0cDSE8mqdKnDEg,51750
 lumibot/backtesting/yahoo_backtesting.py,sha256=LT2524mGlrUSq1YSRnUqGW4-Xcq4USgRv2EhnV_zfs4,502
 lumibot/brokers/__init__.py,sha256=MGWKHeH3mqseYRL7u-KX1Jp2x9EaFO4Ol8sfNSxzu1M,404
 lumibot/brokers/alpaca.py,sha256=VQ17idfqiEFb2JCqqdMGmbvF789L7_PpsCbudiFRzmg,61595
@@ -132,7 +132,7 @@ lumibot/tools/polygon_helper_async.py,sha256=YHDXa9kmkkn8jh7hToY6GP5etyXS9Tj-uky
 lumibot/tools/polygon_helper_polars_optimized.py,sha256=NaIZ-5Av-G2McPEKHyJ-x65W72W_Agnz4lRgvXfQp8c,30415
 lumibot/tools/projectx_helpers.py,sha256=EIemLfbG923T_RBV_i6s6A9xgs7dt0et0oCnhFwdWfA,58299
 lumibot/tools/schwab_helper.py,sha256=CXnYhgsXOIb5MgmIYOp86aLxsBF9oeVrMGrjwl_GEv0,11768
-lumibot/tools/thetadata_helper.py,sha256=RoUIW-JRnuVJzfz42Djhan7cPRa8F8cisQdcMjKpTPM,79366
+lumibot/tools/thetadata_helper.py,sha256=-FJm_NXSBJoyYLcdNQXGytMbmr-wx7F1gItnRnBUWf0,80072
 lumibot/tools/types.py,sha256=x-aQBeC6ZTN2-pUyxyo69Q0j5e0c_swdfe06kfrWSVc,1978
 lumibot/tools/yahoo_helper.py,sha256=htcKKkuktatIckVKfLc_ms0X75mXColysQhrZW244z8,19497
 lumibot/tools/yahoo_helper_polars_optimized.py,sha256=g9xBN-ReHSW4Aj9EMU_OncBXVS1HpfL8LTHit9ZxFY4,7417
@@ -142,7 +142,7 @@ lumibot/traders/trader.py,sha256=KMif3WoZtnSxA0BzoK3kvkTITNELrDFIortx1BYBv8s,967
 lumibot/trading_builtins/__init__.py,sha256=vH2QL5zLjL3slfEV1YW-BvQHtEYLCFkIWTZDfh3y8LE,87
 lumibot/trading_builtins/custom_stream.py,sha256=8_XiPT0JzyXrgnXCXoovGGUrWEfnG4ohIYMPfB_Nook,5264
 lumibot/trading_builtins/safe_list.py,sha256=IIjZOHSiZYK25A4WBts0oJaZNOJDsjZL65MOSHhE3Ig,1975
-lumibot-4.2.0.dist-info/licenses/LICENSE,sha256=fYhGIyxjyNXACgpNQS3xxpxDOaVOWRVxZMCRbsDv8k0,35130
+lumibot-4.2.2.dist-info/licenses/LICENSE,sha256=fYhGIyxjyNXACgpNQS3xxpxDOaVOWRVxZMCRbsDv8k0,35130
 tests/__init__.py,sha256=3-VoT-nAuqMfwufd4ceN6fXaHl_zCfDCSXJOTp1ywYQ,393
 tests/conftest.py,sha256=UBw_2fx7r6TZPKus2b1Qxrzmd4bg8EEBnX1vCHUuSVA,3311
 tests/fixtures.py,sha256=wOHQsh1SGHnXe_PGi6kDWI30CS_Righi7Ig7vwSEKT4,9082
@@ -157,7 +157,7 @@ tests/test_apscheduler_warnings.py,sha256=08lzprPjKq_KAIy-_gMo2zZATpo7VPvmg_qnmS
 tests/test_asset.py,sha256=qk9giu-z3kPoFRXL6Wi0-Ly1mb7YpUNtViuLUMjaEhY,7659
 tests/test_asset_auto_expiry.py,sha256=aa0JVbAIHPKupQ6gMDk5QaDWDXV1xqHMdX513WjwWNQ,13716
 tests/test_auto_market_inference.py,sha256=NFauxzT9ZKDSjrkgHWLgcrLfJMeIj04uWvSKlZVIwqs,1717
-tests/test_backtest_cache_manager.py,sha256=Lkm_xnKX2L-_wS2wGZ57AghPDeZUeoS9BxcuDvwdNnw,4871
+tests/test_backtest_cache_manager.py,sha256=4bZkbv-PZVFRTuKinyLFeAl03tR8rwWV81xHLsvkphI,5120
 tests/test_backtesting_broker.py,sha256=rxZGH5cgiWLmNGdI3k9fti3Fp9IOSohq8xD2E3L2UdY,13194
 tests/test_backtesting_broker_await_close.py,sha256=WbehY7E4Qet3_Mo7lpfgjmhtI9pnJPIt9mkFI15Dzho,7545
 tests/test_backtesting_broker_time_advance.py,sha256=FCv0nKG8BQlEjNft7kmQYm9M2CsLIZ0b7mWCllOHQxc,6378
@@ -234,8 +234,8 @@ tests/test_quiet_logs_requirements.py,sha256=YoUooSVLrFL8TlWPfxEiqxvSj4d8z6-qg58
 tests/test_session_manager.py,sha256=1qygN3aQ2Xe2uh4BMPm0E3V8KXLFNGq5qdL8KkZjef4,11632
 tests/test_strategy_methods.py,sha256=j9Mhr6nnG1fkiVQXnx7gLjzGbeQmwt0UbJr_4plD36o,12539
 tests/test_thetadata_backwards_compat.py,sha256=RzNLhNZNJZ2hPkEDyG-T_4mRRXh5XqavK6r-OjfRASQ,3306
-tests/test_thetadata_helper.py,sha256=lNUhA9VuzCZaGKuN86Zkpq3fi2Qk_R0iv1P5JdkOXNA,57966
-tests/test_thetadata_pandas_verification.py,sha256=MkGIci52W3gez4C6SI4SpcFlrDXmagxikNeOLmXt0yA,6547
+tests/test_thetadata_helper.py,sha256=pcEPu-9kQYp4cn5xmhU1-28DfT-GRu_nUuUMb1xi7nA,58088
+tests/test_thetadata_pandas_verification.py,sha256=MWUecqBY6FGFslWLRo_C5blGbom_unmXCZikAfZXLks,6553
 tests/test_tradier.py,sha256=iCEM2FTxJSzJ2oLNaRqSx05XaX_DCiMzLx1aEYPANko,33280
 tests/test_tradier_data.py,sha256=1jTxDzQtzaC42CQJVXMRMElBwExy1mVci3NFfKjjVH0,13363
 tests/test_tradingfee.py,sha256=2CBJgdU-73Ae4xuys-QkbCtpDTL9hwOUkRnCgLm4OmE,163
@@ -280,7 +280,7 @@ tests/backtest/test_thetadata.py,sha256=xWYfC9C4EhbMDb29qyZWHO3sSWaLIPzzvcMbHCt5
 tests/backtest/test_thetadata_comprehensive.py,sha256=-gN3xLJcJtlB-k4vlaK82DCZDGDmr0LNZZDzn-aN3l4,26120
 tests/backtest/test_thetadata_vs_polygon.py,sha256=dZqsrOx3u3cz-1onIO6o5BDRjI1ey7U9vIkZupfXoig,22831
 tests/backtest/test_yahoo.py,sha256=2FguUTUMC9_A20eqxnZ17rN3tT9n6hyvJHaL98QKpqY,3443
-lumibot-4.2.0.dist-info/METADATA,sha256=fLMcodhb7BQl9rONrvfEQMxIhXx9lo5T-iRftdS2qDk,12092
-lumibot-4.2.0.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
-lumibot-4.2.0.dist-info/top_level.txt,sha256=otUnUjDFVASauEDiTiAzNgMyqQ1B6jjS3QqqP-WSx38,14
-lumibot-4.2.0.dist-info/RECORD,,
+lumibot-4.2.2.dist-info/METADATA,sha256=yjZcnAmbXlQQj4ZEDPZNoWndZHusYMdY8nluNGVQP-0,12092
+lumibot-4.2.2.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
+lumibot-4.2.2.dist-info/top_level.txt,sha256=otUnUjDFVASauEDiTiAzNgMyqQ1B6jjS3QqqP-WSx38,14
+lumibot-4.2.2.dist-info/RECORD,,

tests/test_backtest_cache_manager.py

@@ -67,7 +67,7 @@ def _build_settings(prefix: str = "prod/cache") -> BacktestCacheSettings:
 def test_remote_key_uses_relative_cache_path(tmp_path, monkeypatch):
     cache_root = tmp_path / "cache"
     cache_root.mkdir()
-    local_file = cache_root / "thetadata" / "bars" / "spy.parquet"
+    local_file = cache_root / "thetadata" / "stock" / "minute" / "ohlc" / "stock_SPY_minute_ohlc.parquet"
     local_file.parent.mkdir(parents=True, exist_ok=True)
 
     monkeypatch.setattr(backtest_cache, "LUMIBOT_CACHE_FOLDER", cache_root)
@@ -76,17 +76,17 @@ def test_remote_key_uses_relative_cache_path(tmp_path, monkeypatch):
     manager = BacktestCacheManager(settings, client_factory=lambda settings: StubS3Client())
 
     remote_key = manager.remote_key_for(local_file)
-    assert remote_key == "stage/cache/v3/thetadata/bars/spy.parquet"
+    assert remote_key == "stage/cache/v3/thetadata/stock/minute/ohlc/stock_SPY_minute_ohlc.parquet"
 
 
 def test_ensure_local_file_downloads_from_s3(tmp_path, monkeypatch):
     cache_root = tmp_path / "cache"
     cache_root.mkdir()
-    local_file = cache_root / "thetadata" / "bars" / "spy.parquet"
+    local_file = cache_root / "thetadata" / "stock" / "minute" / "ohlc" / "stock_SPY_minute_ohlc.parquet"
 
     monkeypatch.setattr(backtest_cache, "LUMIBOT_CACHE_FOLDER", cache_root)
 
-    remote_key = "stage/cache/v3/thetadata/bars/spy.parquet"
+    remote_key = "stage/cache/v3/thetadata/stock/minute/ohlc/stock_SPY_minute_ohlc.parquet"
     objects = {("test-bucket", remote_key): b"cached-data"}
 
     stub = StubS3Client(objects)
@@ -101,7 +101,7 @@ def test_ensure_local_file_downloads_from_s3(tmp_path, monkeypatch):
 def test_ensure_local_file_handles_missing_remote(tmp_path, monkeypatch):
     cache_root = tmp_path / "cache"
     cache_root.mkdir()
-    local_file = cache_root / "thetadata" / "bars" / "spy.parquet"
+    local_file = cache_root / "thetadata" / "stock" / "minute" / "ohlc" / "stock_SPY_minute_ohlc.parquet"
 
     monkeypatch.setattr(backtest_cache, "LUMIBOT_CACHE_FOLDER", cache_root)
 
@@ -116,13 +116,13 @@ def test_ensure_local_file_handles_missing_remote(tmp_path, monkeypatch):
 def test_on_local_update_uploads_file(tmp_path, monkeypatch):
     cache_root = tmp_path / "cache"
     cache_root.mkdir()
-    local_file = cache_root / "thetadata" / "bars" / "spy.parquet"
+    local_file = cache_root / "thetadata" / "stock" / "minute" / "ohlc" / "stock_SPY_minute_ohlc.parquet"
     local_file.parent.mkdir(parents=True, exist_ok=True)
     local_file.write_bytes(b"new-data")
 
     monkeypatch.setattr(backtest_cache, "LUMIBOT_CACHE_FOLDER", cache_root)
 
-    remote_key = "stage/cache/v3/thetadata/bars/spy.parquet"
+    remote_key = "stage/cache/v3/thetadata/stock/minute/ohlc/stock_SPY_minute_ohlc.parquet"
     stub = StubS3Client({("test-bucket", remote_key): b"old"})
     manager = BacktestCacheManager(_build_settings(prefix="stage/cache"), client_factory=lambda s: stub)
 

tests/test_thetadata_helper.py

@@ -324,13 +324,13 @@ def test_get_trading_dates():
 def test_build_cache_filename(mocker, tmpdir, datastyle):
     asset = Asset("SPY")
     timespan = "1D"
-    mocker.patch.object(thetadata_helper, "LUMIBOT_CACHE_FOLDER", tmpdir)
-    expected = tmpdir / "thetadata" / f"stock_SPY_1D_{datastyle}.parquet"
+    mocker.patch.object(thetadata_helper, "LUMIBOT_CACHE_FOLDER", str(tmpdir))
+    expected = tmpdir / "thetadata" / "stock" / "1d" / datastyle / f"stock_SPY_1D_{datastyle}.parquet"
     assert thetadata_helper.build_cache_filename(asset, timespan, datastyle) == expected
 
     expire_date = datetime.date(2023, 8, 1)
     option_asset = Asset("SPY", asset_type="option", expiration=expire_date, strike=100, right="CALL")
-    expected = tmpdir / "thetadata" / f"option_SPY_230801_100_CALL_1D_{datastyle}.parquet"
+    expected = tmpdir / "thetadata" / "option" / "1d" / datastyle / f"option_SPY_230801_100_CALL_1D_{datastyle}.parquet"
     assert thetadata_helper.build_cache_filename(option_asset, timespan, datastyle) == expected
 
     # Bad option asset with no expiration
@@ -427,8 +427,8 @@ def test_missing_dates():
     ],
 )
 def test_update_cache(mocker, tmpdir, df_all, df_cached, datastyle):
-    mocker.patch.object(thetadata_helper, "LUMIBOT_CACHE_FOLDER", tmpdir)
-    cache_file = Path(tmpdir / "thetadata" / f"stock_SPY_1D_{datastyle}.parquet")
+    mocker.patch.object(thetadata_helper, "LUMIBOT_CACHE_FOLDER", str(tmpdir))
+    cache_file = thetadata_helper.build_cache_filename(Asset("SPY"), "1D", datastyle)
     
     # Empty DataFrame of df_all, don't write cache file
     thetadata_helper.update_cache(cache_file, df_all, df_cached)
@@ -550,8 +550,9 @@ def test_get_price_data_invokes_remote_cache_manager(tmp_path, monkeypatch):
 )
 def test_load_data_from_cache(mocker, tmpdir, df_cached, datastyle):
     # Setup some basics
-    mocker.patch.object(thetadata_helper, "LUMIBOT_CACHE_FOLDER", tmpdir)
-    cache_file = Path(tmpdir / "thetadata" / f"stock_SPY_1D_{datastyle}.parquet")
+    mocker.patch.object(thetadata_helper, "LUMIBOT_CACHE_FOLDER", str(tmpdir))
+    asset = Asset("SPY")
+    cache_file = thetadata_helper.build_cache_filename(asset, "1D", datastyle)
 
     # No cache file should return None (not raise)
     assert thetadata_helper.load_cache(cache_file) is None
@@ -1371,8 +1372,8 @@ class TestThetaDataChainsCaching:
 
         # CLEAR CACHE to ensure first call downloads fresh data
         # This prevents cache pollution from previous tests in the suite
-        # Chains are stored in: LUMIBOT_CACHE_FOLDER / "thetadata" / "option_chains"
-        chain_folder = Path(LUMIBOT_CACHE_FOLDER) / "thetadata" / "option_chains"
+        # Chains are stored in: LUMIBOT_CACHE_FOLDER / "thetadata" / "option" / "option_chains"
+        chain_folder = Path(LUMIBOT_CACHE_FOLDER) / "thetadata" / "option" / "option_chains"
         if chain_folder.exists():
             # Delete all AAPL chain cache files
             for cache_file in chain_folder.glob("AAPL_*.parquet"):

tests/test_thetadata_pandas_verification.py

@@ -44,7 +44,7 @@ def count_cache_files():
     cache_dir = get_cache_dir()
     if not cache_dir.exists():
         return 0
-    return len(list(cache_dir.glob("*.parquet")))
+    return sum(1 for _ in cache_dir.rglob("*.parquet"))
 
 
 class WeeklyMomentumOptionsStrategy(Strategy):