chore: add weather pipeline example

examples/02_demonstration/weather/custom_types.midas

@@ -0,0 +1,65 @@
+predicate in_range(min: float, max: float)(v: float) = min <= v & v <= max
+predicate is_percentage = in_range(0.0, 100.0)
+
+type Celsius = float
+type Kelvin = float where _ >= 0
+type Hectopascal = float
+
+type Temperature = Celsius where in_range(-30.0, 100.0)
+type Pressure = Hectopascal where in_range(800.0, 1100.0)(_)
+type Humidity = float where is_percentage(_)
+type HeatIndex = float
+
+type StationID = str where len(_) == 3 & _.isupper()
+
+type Mean[T <: float] = float
+
+extend Celsius {
+    def __add__: fn(Celsius, /) -> Celsius
+    def __sub__: fn(Celsius, /) -> Celsius
+}
+
+extend Kelvin {
+    def __add__: fn(Kelvin, /) -> Kelvin
+}
+
+extend Hectopascal {
+    def __add__: fn(Hectopascal, /) -> Hectopascal
+    def __sub__: fn(Hectopascal, /) -> Hectopascal
+}
+
+alias RawData = Frame[
+    station_id: str,
+    timestamp: str,
+    temperature: float,
+    pressure: float,
+    humidity: float,
+]
+
+alias Data = Frame[
+    station_id: StationID,
+    timestamp: object,
+    temperature: Temperature,
+    pressure: Pressure,
+    humidity: Humidity,
+]
+
+alias DataWithHI = Frame[
+    station_id: StationID,
+    timestamp: object,
+    temperature: Temperature,
+    pressure: Pressure,
+    humidity: Humidity,
+    heat_index: HeatIndex,
+]
+
+alias DailyAverages = Frame[
+    timestamp: object,
+    temperature: Mean[Temperature],
+    pressure: Mean[Pressure],
+    humidity: Mean[Humidity],
+    heat_index: Mean[HeatIndex],
+]
+
+predicate limit_amplitude(max_amp: float)(ls: list[float]) = max(ls) - min(ls) <= max_amp
+type LowAmplitudeWave = list[float where _ >= 1] where limit_amplitude(10)(_)

examples/02_demonstration/weather/gen_data.py

@@ -0,0 +1,43 @@
+import datetime
+import random
+
+import pandas as pd
+
+stations = ["SIO", "AIG", "ZER"]
+
+start_ts = datetime.datetime(2026, 1, 1)
+end_ts = datetime.datetime(2027, 1, 1)
+delta = end_ts - start_ts
+
+min_temp, max_temp = -30.0, 100.0
+min_pres, max_pres = 800.0, 1100.0
+min_hum, max_hum = 0.0, 1.0
+
+N = 3000
+
+rows: list[tuple[str, datetime.datetime, float, float, float]] = []
+
+for _ in range(N):
+    ts = random.random() * delta + start_ts
+    rows.append(
+        (
+            random.choice(stations),
+            ts,
+            random.random() * (max_temp - min_temp) + min_temp,
+            random.random() * (max_pres - min_pres) + min_pres,
+            random.random() * (max_hum - min_hum) + min_hum,
+        )
+    )
+
+df = pd.DataFrame(
+    rows,
+    columns=[
+        "station_id",
+        "timestamp",
+        "temperature",
+        "pressure",
+        "humidity",
+    ],
+)
+df = df.sort_values(by=["timestamp", "station_id"])
+df.to_csv("data.csv", index=False)

examples/02_demonstration/weather/pipeline.py

@@ -0,0 +1,69 @@
+from pathlib import Path
+
+import matplotlib.pyplot as plt
+import pandas as pd
+from custom_types import DailyAverages, Data, DataWithHI, HeatIndex, RawData
+
+from midas.typing import Column, cast, unsafe_cast
+
+
+def load_data(path: Path) -> RawData:
+    return cast(RawData, pd.read_csv(path))
+
+
+def convert_data(raw_df: RawData) -> Data:
+    new_df = raw_df.copy()
+    new_df["timestamp"] = cast(
+        Column[object],
+        pd.to_datetime(new_df["timestamp"]),
+    )
+    return cast(Data, new_df)
+
+
+def compute_heat_index(df: Data):
+    df["heat_index"] = cast(
+        Column[HeatIndex],
+        (
+            df["temperature"] * 2.0
+            + df["humidity"] * 10.0
+            - df["temperature"] * df["humidity"] * 0.2
+        ),
+    )
+    return df
+
+
+def daily_avg(df: DataWithHI):
+    return cast(
+        DailyAverages,
+        df.groupby(
+            by=[
+                df["station_id"],
+                df["timestamp"].dt.day.rename("day"),
+            ],
+        )
+        .mean()
+        .sort_values(by="timestamp"),
+    )
+
+
+def plot(df: DailyAverages):
+    stations = unsafe_cast(list[str], list(df.index.get_level_values(0).unique()))
+    for station in stations:
+        sub_df = unsafe_cast(DailyAverages, df.loc[station])
+        # plt.plot(sub_df["timestamp"], sub_df["temperature"])
+        plt.plot(sub_df["timestamp"], sub_df["heat_index"])
+    plt.show()
+
+
+def main():
+    raw_df: RawData = load_data(Path("data.csv"))
+    df: Data = convert_data(raw_df)
+
+    with_hi = compute_heat_index(df)
+    dailies = daily_avg(with_hi)
+    print(dailies)
+    plot(dailies)
+
+
+if __name__ == "__main__":
+    main()