add preprocessing steps

src/delta_barth/analysis/forecast.py

@@ -1,19 +1,34 @@
 from __future__ import annotations
 
 import dataclasses as dc
+from collections.abc import Mapping, Set
 from typing import TYPE_CHECKING
 
 import pandas as pd
 from sklearn.metrics import mean_squared_error
 from xgboost import XGBRegressor
 
+from delta_barth.analysis import parse
+from delta_barth.constants import COL_MAP_SALES_PROGNOSIS, FEATURES_SALES_PROGNOSIS
 from delta_barth.types import CustomerDataSalesForecast, FcErrorCodes
 
 if TYPE_CHECKING:
+    from delta_barth.api.common import SalesPrognosisResponse
     from delta_barth.types import FcResult
 
 # TODO check pandera for DataFrame validation
 
+
+def parse_api_resp_to_df(
+    resp: SalesPrognosisResponse,
+) -> pd.DataFrame:
+    if resp.error is not None:
+        raise ValueError("Response contains error code. Parsing aborted.")
+    data = resp.model_dump()["daten"]
+
+    return pd.DataFrame(data)
+
+
 # ------------------------------------------------------------------------------
 # Input:
 #   DataFrame df mit Columns f_umsatz_fakt, firmen, art, v_warengrp
@@ -28,8 +43,22 @@ if TYPE_CHECKING:
 
 # Prognose Umsatz je Firma
 
+
 # TODO: check usage of separate exception and handle it in API function
 # TODO set min number of data points as constant, not parameter
+def preprocess_sales_per_customer(
+    resp: SalesPrognosisResponse,
+    feature_map: Mapping[str, str],
+    target_features: Set[str],
+) -> pd.DataFrame:
+    df = parse_api_resp_to_df(resp)
+    df = parse.preprocess_features(
+        df,
+        feature_map=feature_map,
+        target_features=target_features,
+    )
+
+    return df
 
 
 def sales_per_customer(

src/delta_barth/analysis/parse.py

@@ -9,9 +9,9 @@ if TYPE_CHECKING:
     import pandas as pd
 
 
-def check_needed_features(
+def _check_needed_features(
     data: pd.DataFrame,
-    features: Set,
+    features: Set[str],
 ) -> None:
     data_feats = set(data.columns)
     missing_features = features - data_feats
@@ -23,10 +23,11 @@ def check_needed_features(
         )
 
 
-def map_features_to_targets(
+def _map_features_to_targets(
     data: pd.DataFrame,
     feature_map: Mapping[str, str],
 ) -> pd.DataFrame:
+    data = data.copy()
     data_feats = data.columns
     mapped_feats: list[str] = []
 
@@ -39,3 +40,14 @@ def map_features_to_targets(
     data.columns = mapped_feats
 
     return data
+
+
+def preprocess_features(
+    data: pd.DataFrame,
+    feature_map: Mapping[str, str],
+    target_features: Set[str],
+) -> pd.DataFrame:
+    data = _map_features_to_targets(data, feature_map)
+    _check_needed_features(data, target_features)
+
+    return data

tests/analysis/test_forecast.py

@@ -1,3 +1,5 @@
+import pytest
+
 from delta_barth.analysis import forecast as fc
 
 
@@ -15,3 +17,44 @@ def test_sales_per_customer_too_few_data_points(sales_data):
 
     assert err == 1
     assert res is None
+
+
+def test_parse_api_resp_to_df(exmpl_api_sales_prognosis_resp):
+    resp = exmpl_api_sales_prognosis_resp
+    df = fc.parse_api_resp_to_df(resp)
+    features = set(
+        (
+            "artikelId",
+            "warengruppeId",
+            "firmaId",
+            "betrag",
+            "menge",
+            "buchungsDatum",
+        )
+    )
+    assert all(col in features for col in df.columns)
+
+
+def test_preprocess_sales_per_customer(exmpl_api_sales_prognosis_resp):
+    resp = exmpl_api_sales_prognosis_resp
+    feat_mapping: dict[str, str] = {
+        "artikelId": "artikel_refid",
+        "firmaId": "firma_refid",
+        "betrag": "betrag",
+        "menge": "menge",
+        "buchungsDatum": "buchungs_datum",
+    }
+    target_features: frozenset[str] = frozenset(
+        (
+            "firma_refid",
+            "betrag",
+            "buchungs_datum",
+        )
+    )
+    df = fc.preprocess_sales_per_customer(
+        resp,
+        feature_map=feat_mapping,
+        target_features=target_features,
+    )
+    assert len(df.columns) == 5
+    assert any(feat not in df.columns for feat in feat_mapping.keys())

tests/analysis/test_parse.py

@@ -1,7 +1,7 @@
 import pandas as pd
 import pytest
 
-from delta_barth.analysis import parse
+from delta_barth.analysis import forecast, parse
 from delta_barth.errors import FeaturesMissingError
 
 
@@ -10,12 +10,12 @@ def test_check_needed_features():
     data = pd.DataFrame(
         data=[[1, 2, 3, 4, 5]], columns=["feat1", "feat2", "feat3", "feat4", "feat5"]
     )
-    parse.check_needed_features(data, target_features)
+    parse._check_needed_features(data, target_features)
     data = pd.DataFrame(
         data=[[1, 2, 3, 4, 5]], columns=["featX", "feat2", "feat3", "feat4", "feat5"]
     )
     with pytest.raises(FeaturesMissingError):
-        parse.check_needed_features(data, target_features)
+        parse._check_needed_features(data, target_features)
 
 
 def test_map_features_to_targets():
@@ -23,7 +23,7 @@ def test_map_features_to_targets():
     data = pd.DataFrame(
         data=[[1, 2, 3, 4, 5]], columns=["feat1", "feat2", "feat3", "feat4", "feat5"]
     )
-    data = parse.map_features_to_targets(data, feature_map)
+    data = parse._map_features_to_targets(data, feature_map)
     assert "feat10" in data.columns
     assert "feat20" in data.columns
     assert "feat50" in data.columns
@@ -32,3 +32,28 @@ def test_map_features_to_targets():
     assert "feat1" not in data.columns
     assert "feat2" not in data.columns
     assert "feat5" not in data.columns
+
+
+def test_preprocess_features(exmpl_api_sales_prognosis_resp):
+    resp = exmpl_api_sales_prognosis_resp
+    df = forecast.parse_api_resp_to_df(resp)
+    feat_mapping: dict[str, str] = {
+        "artikelId": "artikel_refid",
+        "firmaId": "firma_refid",
+        "betrag": "betrag",
+        "menge": "menge",
+        "buchungsDatum": "buchungs_datum",
+    }
+    target_features: frozenset[str] = frozenset(
+        (
+            "firma_refid",
+            "betrag",
+            "buchungs_datum",
+        )
+    )
+
+    assert all(feat in df.columns for feat in feat_mapping.keys())
+    data = parse.preprocess_features(df, feat_mapping, target_features)
+    assert len(data.columns) == len(df.columns)
+    assert (data.columns != df.columns).any()
+    assert any(feat not in data.columns for feat in feat_mapping.keys())

tests/conftest.py

@@ -1,10 +1,16 @@
+from __future__ import annotations
+
+import pickle
 import tomllib
 from pathlib import Path
-from typing import Any, cast
+from typing import TYPE_CHECKING, Any, cast
 
 import pandas as pd
 import pytest
 
+if TYPE_CHECKING:
+    from delta_barth.api.common import SalesPrognosisResponse
+
 
 @pytest.fixture(scope="session")
 def credentials() -> dict[str, str]:
@@ -56,3 +62,16 @@ def sales_data() -> pd.DataFrame:
     data["buchungs_datum"] = pd.to_datetime(data["buchungs_datum"])
 
     return data
+
+
+@pytest.fixture(scope="session")
+def exmpl_api_sales_prognosis_resp() -> SalesPrognosisResponse:
+    pwd = Path.cwd()
+    assert "barth" in pwd.parent.name.lower(), "not in project root directory"
+    data_pth = pwd / "./tests/_test_data/exmp_sales_prognosis_resp.pkl"
+    assert data_pth.exists(), "file to API sales data not found"
+
+    with open(data_pth, "rb") as file:
+        data = cast("SalesPrognosisResponse", pickle.load(file))
+
+    return data