The demo app shows how easy it is to connect to OpenWeatherMap's API using Streamlit's experimental_connection. This feature streamlines the handling of secrets and other repetitive code often reused to connect to the OpenWeatherMap API. See the app here.
To run it locally: in a .streamlit/secrets.toml file, you'll need to provide the api key as follows:
owm_api_key = "____"
This repo contains 2 types of connections created from Streamlit's ExperimentalBaseConnection. They access the OpenWeatherMap API using the PyOWM package.
The first connection returns an OWM object. Example usage:
from openweathermap_connection import OpenWeatherMapConnection
conn = st.experimental_connection(name='example_conn', type=OpenWeatherMapConnection)
city_cur = conn.city_id_registry()
From there you can access the following entry points that pyowm provides:
- agro_manager()
- airpollution_manager()
- alert_manager()
- city_id_registry()
- geocoding_manager()
- stations_manager()
- tile_manager()
- uvindex_manager()
- weather_manager()
The second connection returns an entry point to OneCall. Example usage:
from openweathermap_connection import OpenWeatherMapOneCallConnection
conn = st.experimental_connection('weather', lat=40.1, lon=-70.5, units='imperial', type=OpenWeatherMapOneCallConnection, ttl=3600)
temperature_data = conn.get_temperature_df()
current_temperature = conn.get_current()
This class lets you retrieve weather data with the pyowm weather manager.
get_temperature_df is a custom method that takes the hourly and daily forecasts and combines them into a single pandas dataframe with datetime and temperature.