Skip to content

Weather#

smart_control.simulator.weather_controller #

Controls ambient temperature in simulator.

BaseWeatherController #

Represents the weather on any specific time.

get_current_temp abstractmethod #

get_current_temp(timestamp: Timestamp) -> float

Gets outside temp at specified timestamp.

ReplayWeatherController #

ReplayWeatherController(
    local_weather_path: str = WEATHER_CSV_FILEPATH,
    convection_coefficient: float = 12.0,
    humidity_column: str = "Humidity",
)

Bases: BaseWeatherController

Weather controller that interplolates real weather from past observations.

Attributes:

Name Type Description
local_weather_path

Path to local weather CSV file.
weather_df

Pandas dataframe of historical weather data.
convection_coefficient

Air convection coefficient (W/m2/K).
humidity_column

Column name of the humidity in the weather CSV file.

csv_filepath property #

csv_filepath: str

Alias for the local weather CSV file path.

humidities property #

humidities: Series

Returns the humidities of the weather data.

max_time property #

max_time: Timestamp

Latest timestamp in the weather data.

min_time property #

min_time: Timestamp

Earliest timestamp in the weather data.

temps_f property #

temps_f: Series

Returns the temperatures in Fahrenheit of the weather data.

times_in_seconds property #

times_in_seconds: Index

Returns the timestamps of the weather data, as seconds since epoch.

get_current_humidity #

get_current_humidity(timestamp: Timestamp) -> float

For a given timestamp, returns the current humidity level in percent.

get_current_temp #

get_current_temp(timestamp: Timestamp) -> float

For a given timestamp, returns the current temperature in Kelvin.

read_weather_csv #

read_weather_csv(csv_filepath: str) -> pd.DataFrame

Loads time series weather data from the specified CSV file.

The CSV file is expected to have at least the following columns:

  • Time: the time, as a string, in the format: %Y%m%d-%H%M (e.g. 20230701-0000). Assumed to be in UTC.
  • TempF: the temperature in Fahrenheit at the specified time.
  • Humidity: the relative humidity in percent at the specified time (0 to 100).

Coerces the times to UTC. Updates the index to be seconds since epoch.

Parameters:

Name Type Description Default
csv_filepath str

Path to local weather CSV file.

 required

Returns:

Type Description
DataFrame

Pandas dataframe of weather data.

WeatherController #

WeatherController(
    default_low_temp: float,
    default_high_temp: float,
    special_days: Optional[Mapping[int, TemperatureBounds]] = None,
    convection_coefficient: float = 12.0,
)

Bases: BaseWeatherController

Represents the weather on any given day.

Weather operates as a sinusoid: low at midnight and high at noon.

Attributes:

Name Type Description
default_low_temp

Default low temperature in K at midnight.
default_high_temp

Default high temperature in K at noon.
special_days

Map of day of year (1-365) to 2-tuple (low_temp, high_temp).
convection_coefficient

Air convection coefficient (W/m2/K).

get_air_convection_coefficient #

get_air_convection_coefficient(timestamp: Timestamp) -> float

Returns the convection coefficient (W/m2/K) based on the current wind.

Parameters:

Name Type Description Default
timestamp Timestamp

Pandas timestamp to get convection coefficient for.

 required

get_current_temp #

get_current_temp(timestamp: Timestamp) -> float

Returns current temperature in K.

Parameters:

Name Type Description Default
timestamp Timestamp

Pandas timestamp to get temperature for.

 required

seconds_to_rads #

seconds_to_rads(seconds_in_day: int) -> float

Returns radians corresponding to number of second in the day.

Parameters:

Name Type Description Default
seconds_in_day int

Seconds that have passed so far in the day.

 required

get_replay_temperatures #

get_replay_temperatures(
    observation_responses: Sequence[ObservationResponse],
) -> Mapping[str, float]

Returns temperature replays from past observations.

Parameters:

Name Type Description Default
observation_responses Sequence[ObservationResponse]

array of observations to extract weather from

 required

Returns: map from timestamp to temp