first sktime (NaiveForecaster) model

models/st-naive/Dockerfile

@@ -0,0 +1,16 @@
+FROM python:3.11.6-slim-bookworm
+
+WORKDIR /usr/local/app
+
+COPY ./requirements.txt /usr/local/app/requirements.txt
+RUN pip install --no-cache-dir --upgrade pip
+RUN pip install --no-cache-dir -r /usr/local/app/requirements.txt
+
+# Polars requires AVX2 CPU instructions, which are not available on the server
+RUN pip uninstall -y polars
+RUN pip install --no-cache-dir polars-lts-cpu
+
+COPY ./src /usr/local/app/src
+
+EXPOSE 3000
+CMD ["uvicorn", "src.main:app", "--host", "0.0.0.0", "--port", "3000"]

models/st-naive/requirements.txt

@@ -0,0 +1,2 @@
+enfobench>=0.3.5,<0.4.0
+sktime==0.24.1

models/st-naive/src/main.py

@@ -0,0 +1,48 @@
+import pandas as pd
+from sktime.forecasting.naive import NaiveForecaster
+
+from enfobench import AuthorInfo, ForecasterType, ModelInfo
+from enfobench.evaluation.server import server_factory
+from enfobench.evaluation.utils import create_forecast_index
+
+
+class NaiveForecasterModel:
+    def info(self) -> ModelInfo:
+        return ModelInfo(
+            name="Sktime.NaiveForecaster.Mean",
+            authors=[AuthorInfo(name="Attila Balint", email="[email protected]")],
+            type=ForecasterType.point,
+            params={
+                "strategy": "mean",
+            },
+        )
+
+    def forecast(
+        self,
+        horizon: int,
+        history: pd.DataFrame,
+        past_covariates: pd.DataFrame | None = None,
+        future_covariates: pd.DataFrame | None = None,
+        level: list[int] | None = None,
+        **kwargs,
+    ) -> pd.DataFrame:
+        # Fill missing values
+        y = history.y.fillna(history.y.mean())
+
+        # Create model
+        model = NaiveForecaster(strategy="mean")
+
+        # Make forecast
+        index = create_forecast_index(history=history, horizon=horizon)
+        pred: pd.Series = model.fit_predict(y, fh=index, **kwargs)
+
+        # Postprocess forecast
+        forecast = pred.to_frame("yhat").fillna(y.mean())
+        return forecast
+
+
+# Instantiate your model
+model = NaiveForecasterModel()
+
+# Create a forecast server by passing in your model
+app = server_factory(model)