Move decomposition module into methods sub-package

- move the module
- unify the corresponding tests in `tests.forecasts.methods` sub-package
- make all `predict()` and the `stl()` function round results
- streamline documentation

setup.cfg

@@ -147,7 +147,7 @@ per-file-ignores =
     src/urban_meal_delivery/db/restaurants.py:
         # The module is not too complex.
         WPS232,
-    src/urban_meal_delivery/forecasts/decomposition.py:
+    src/urban_meal_delivery/forecasts/methods/decomposition.py:
         # The module is not too complex.
         WPS232,
     src/urban_meal_delivery/forecasts/timify.py:

src/urban_meal_delivery/forecasts/__init__.py

@@ -1,5 +1,4 @@
 """Demand forecasting utilities."""
 
-from urban_meal_delivery.forecasts import decomposition
 from urban_meal_delivery.forecasts import methods
 from urban_meal_delivery.forecasts import timify

src/urban_meal_delivery/forecasts/methods/__init__.py

@@ -1,4 +1,5 @@
 """Various forecasting methods implemented as functions."""
 
 from urban_meal_delivery.forecasts.methods import arima
+from urban_meal_delivery.forecasts.methods import decomposition
 from urban_meal_delivery.forecasts.methods import ets

src/urban_meal_delivery/forecasts/methods/arima.py

@@ -14,7 +14,7 @@ def predict(
 ) -> pd.DataFrame:
     """Predict with an automatically chosen ARIMA model.
 
-    Note: The function does not check if the `forecast` interval
+    Note: The function does not check if the `forecast_interval`
     extends the `training_ts`'s interval without a gap!
 
     Args:
@@ -65,7 +65,7 @@ def predict(
     forecasts = pandas2ri.rpy2py(result)
     forecasts.index = forecast_interval
 
-    return forecasts.rename(
+    return forecasts.round(5).rename(
         columns={
             'Point Forecast': 'prediction',
             'Lo 80': 'low80',

src/urban_meal_delivery/forecasts/methods/decomposition.py

@@ -169,10 +169,13 @@ def stl(  # noqa:C901,WPS210,WPS211,WPS231
 
     # Unpack the result to a `pd.DataFrame`.
     result = pandas2ri.rpy2py(result[0])
-    result = {
-        'seasonal': pd.Series(result[:, 0], index=time_series.index),
-        'trend': pd.Series(result[:, 1], index=time_series.index),
-        'residual': pd.Series(result[:, 2], index=time_series.index),
-    }
+    result = pd.DataFrame(
+        data={
+            'seasonal': result[:, 0],
+            'trend': result[:, 1],
+            'residual': result[:, 2],
+        },
+        index=time_series.index,
+    )
 
-    return pd.DataFrame(result)
+    return result.round(5)

src/urban_meal_delivery/forecasts/methods/ets.py

@@ -14,7 +14,7 @@ def predict(
 ) -> pd.DataFrame:
     """Predict with an automatically calibrated ETS model.
 
-    Note: The function does not check if the `forecast` interval
+    Note: The function does not check if the `forecast_interval`
     extends the `training_ts`'s interval without a gap!
 
     Args:
@@ -66,7 +66,7 @@ def predict(
     forecasts = pandas2ri.rpy2py(result)
     forecasts.index = forecast_interval
 
-    return forecasts.rename(
+    return forecasts.round(5).rename(
         columns={
             'Point Forecast': 'prediction',
             'Lo 80': 'low80',

tests/forecasts/methods/__init__.py

@@ -0,0 +1 @@
+"""Tests for the `urban_meal_delivery.forecasts.methods` sub-package."""

tests/forecasts/methods/test_decomposition.py

@@ -7,7 +7,7 @@ import pytest
 
 from tests.forecasts.conftest import NS
 from tests.forecasts.conftest import VERTICAL_FREQUENCY
-from urban_meal_delivery.forecasts import decomposition
+from urban_meal_delivery.forecasts.methods import decomposition
 
 
 class TestInvalidArguments:

tests/forecasts/methods/test_ts_methods.py

@@ -1,4 +1,7 @@
-"""Test the `arima.predict()` and `ets.predict()` functions."""
+"""Test the `arima.predict()` and `ets.predict()` functions.
+
+We consider both "classical" time series prediction models.
+"""
 
 import datetime as dt