Dev

.gitignore

@@ -4,8 +4,8 @@
 src/moddata/data/archive.zip
 .DS_Store
 src/moddata/data/btc.parquet
-src/moddata/data/stooq_data/*
-src/moddata/data/regime_model_data/*
+src/moddata/data/stooq_data
+src/moddata/data/regime_model_data
 sandbox.py
 .coverage
 moddata.egg-info

src/moddata/_utils.py

@@ -17,7 +17,8 @@
     "spx_1901-2025",
     "sunspots",
     "geomagnetic_activity",
-    "world_bank_oil_gold_monthly_prices"
+    "world_bank_oil_gold_monthly_prices",
+    "eurpln_regime"
 ]
 
 
@@ -92,6 +93,11 @@ def _load_spx_1901_to_2025() -> pd.DataFrame:
         resources.files('moddata.data').joinpath('spx_1901-2025.parquet')  # noqa
     ))
 
+def _load_eurpln_regime() -> pd.DataFrame:
+    return pd.read_parquet(str(
+        resources.files('moddata.data').joinpath('eurpln_regime.parquet')  # noqa
+    ))
+
 
 def load_data(dataset: Dataset) -> pd.DataFrame | None:
     if dataset == "bankchurn":
@@ -108,4 +114,6 @@ def load_data(dataset: Dataset) -> pd.DataFrame | None:
         return _load_world_bank_oil_gold_monthly_prices()
     if dataset == "spx_1901-2025":
         return _load_spx_1901_to_2025()
+    if dataset == "eurpln_regime":
+        return _load_eurpln_regime()
     raise ValueError(f"Encountered invalid dataset name: {dataset}")

src/moddata/common/aliases.py

@@ -0,0 +1,4 @@
+from typing import Literal, TypeAlias
+
+
+JoinStyle: TypeAlias = Literal["outer", "inner"]

src/moddata/extractor/stooq_data_extractor.py

@@ -0,0 +1,32 @@
+from pathlib import Path
+from typing import Final
+
+import pandas as pd
+
+
+class StooqDataExtractor:
+
+    def __init__(
+            self,
+            data_folder: Path,
+            files: list[str]
+    ):
+        self._data_folder: Final[Path] = data_folder
+        self._files: Final[list[str]] = files
+
+    def _retrieve_ticker(self, file: str) -> str:
+        return file.split(".")[0].split("_")[0]
+
+    def _extract_file(self, file: str) -> pd.DataFrame:
+        data = pd.read_csv(self._data_folder / file, sep=",", decimal=".",)
+        data.columns = [c.lower() for c in data.columns]
+        data["date"] = pd.to_datetime(data["date"])
+        data = data.set_index("date")
+        data.columns = pd.MultiIndex.from_product([
+            [self._retrieve_ticker(file)],
+            list(data.columns),
+        ], names=["ticker", "variable"])
+        return data
+
+    def extract(self) -> dict[str, pd.DataFrame]:
+        return {file: self._extract_file(file) for file in self._files}

src/moddata/pipeline/eurpln_regime_modeling_data_pipeline.py

@@ -0,0 +1,66 @@
+from datetime import date
+from pathlib import Path
+
+import pandas as pd
+
+from moddata.extractor.stooq_data_extractor import StooqDataExtractor
+from moddata.extractor.cboe_vix_data_extractor import CboeVIXDataExtractor
+from moddata.transformer.merge_stooq_data_transformer import (
+    MergeStooqDataTransformer)
+
+
+class EURPLNRegimeModelingDataPipeline:
+
+    def __init__(
+            self,
+            start: date = date(2016, 1, 1),
+            end: date = date(2025, 12, 31)
+    ):
+        self._start, self._end = start, end
+        self._stooq_extractor = StooqDataExtractor(
+            data_folder=(
+                    Path(__file__).parent.parent /
+                    "data" /
+                    "regime_model_data"
+            ),
+            files=[
+                "10ydey_b_d.csv",
+                "10yply_b_d.csv",
+                "10yusy_b_d.csv",
+                "^uslc_d.csv",
+                "eurczk_d.csv",
+                "eurhuf_d.csv",
+                "eurpln_d.csv",
+                "eurusd_d.csv",
+                "usdpln_d.csv"
+            ]
+        )
+        self._stooq_transformer = MergeStooqDataTransformer()
+        self._vix_extractor = CboeVIXDataExtractor()
+
+    def run(self):
+        stooq_data = self._stooq_extractor.extract()
+        stooq_data = self._stooq_transformer.transform(stooq_data)
+
+        vix_data = self._vix_extractor.extract()
+
+        data = pd.merge(
+            left=stooq_data,
+            right=vix_data,
+            how="outer",
+            left_index=True,
+            right_index=True
+        )
+        data = data[f"{self._start:%Y-%m-%d}":f"{self._end:%Y-%m-%d}"]
+        data = data.ffill().bfill()
+
+        data.to_parquet(
+            Path(__file__).parent.parent /
+            "data" /
+            "eurpln_regime.parquet",
+            engine="pyarrow"
+        )
+
+
+if __name__ == "__main__":
+    EURPLNRegimeModelingDataPipeline().run()

src/moddata/transformer/merge_stooq_data_transformer.py

@@ -0,0 +1,27 @@
+from functools import reduce
+
+import pandas as pd
+
+from moddata.common.aliases import JoinStyle
+
+
+class MergeStooqDataTransformer:
+
+    def __init__(self, join_style: JoinStyle = "outer"):
+        self._join_style: JoinStyle = join_style
+
+    def transform(self, data: dict[str, pd.DataFrame]) -> pd.DataFrame:
+        data: pd.DataFrame = reduce(
+            lambda l_, r_: pd.merge(
+                left=l_,
+                right=r_,
+                left_index=True,
+                right_index=True,
+                how=self._join_style # noqa
+            ),
+            data.values()
+        )
+        data = data.reset_index(drop=False).set_index("date")
+        if self._join_style == "outer":
+            data = data.ffill().bfill()
+        return data