Skip to main content

Documentation Index

Fetch the complete documentation index at: https://mintlify.com/elenacarino-max/mas-climapp/llms.txt

Use this file to discover all available pages before exploring further.

Every weather record that flows through ClimApp is automatically evaluated against a set of meteorological risk thresholds before it is saved or returned to the browser. This evaluation is handled by AlertService in services/alert_service.py and produces a list of human-readable alert codes that the frontend can display directly.

​
How alert evaluation works

AlertService.evaluar_alertas() accepts a single argument: a normalized weather dictionary produced by NormalizerService. It reads four keys from that dictionary — temperatura, viento, lluvia, and humedad — and compares each value against the thresholds below. Alert codes from different variables can coexist in the same response; for example, a record can simultaneously carry NARANJA_CALOR and NARANJA_VIENTO. If the data dictionary fails the validate_weather_data() check in utils/validators.py, the method returns an empty list immediately and skips all threshold comparisons. Likewise, if any value cannot be cast to float, the method returns an empty list rather than raising an exception.

​
Alert codes and thresholds

The table below lists every alert code that AlertService can emit, in priority order within each variable group.
Alert codeVariableCondition
ROJA_CALORTemperature≥ 40 °C
NARANJA_CALORTemperature≥ 35 °C and < 40 °C
ROJA_FRIOTemperature≤ −5 °C
NARANJA_FRIOTemperature≤ 0 °C and > −5 °C
ROJA_VIENTOWind speed> 70 km/h
NARANJA_VIENTOWind speed> 40 km/h and ≤ 70 km/h
ROJA_LLUVIARainfall> 30 mm
NARANJA_LLUVIARainfall> 10 mm and ≤ 30 mm
NARANJA_HUMEDADHumidity≥ 90%
VERDE—No other condition met
Temperature uses elif branches, so only one temperature code is ever emitted per record. Wind and rainfall behave the same way. Humidity has only a single threshold — there is no ROJA_HUMEDAD level. All four variable checks run independently, which means the alertas list can contain at most one temperature code, one wind code, one rainfall code, and one humidity code at the same time.

​
Example API response

The alertas field is included directly in the JSON object returned by GET /api/clima. The example below shows a hot-day scenario where temperature breaches the 35 °C orange threshold but does not reach 40 °C: 
{
  "ciudad": "Madrid-Cuatro Vientos",
  "estacion": "Madrid-Cuatro Vientos",
  "fecha": "2026-05-07T14:00:00",
  "temperatura": 37.5,
  "humedad": 45,
  "viento": 25.0,
  "presion": 1012.0,
  "lluvia": 0.0,
  "alertas": ["NARANJA_CALOR"],
  "fuente": "aemet"
}
A record with no risk conditions returns: 
{
  "ciudad": "Madrid-Retiro",
  "estacion": "Madrid-Retiro",
  "fecha": "2026-05-07T09:00:00",
  "temperatura": 18.2,
  "humedad": 62,
  "viento": 12.0,
  "presion": 1018.0,
  "lluvia": 0.0,
  "alertas": ["VERDE"],
  "fuente": "aemet"
}

​
Validation before evaluation

validate_weather_data() runs as the first line of evaluar_alertas(). Invalid data — such as a record missing required fields or containing out-of-range values — causes the method to return [] rather than emitting misleading alert codes. This ensures that only structurally sound, physically plausible readings are ever classified.
alert_service.py
def evaluar_alertas(self, registro_normalizado: Dict[str, Any]) -> List[str]:
    if not validate_weather_data(registro_normalizado):
        return []
    # threshold checks follow ...
VERDE is not a default fallback applied blindly — it is only appended when the alertas list is still empty after all four variable checks have completed. A record carrying VERDE means every measured variable is within safe operating ranges.

Build docs developers (and LLMs) love

Get started for freeTalk to us