Welcome to Solcast

A simple Python SDK that wraps Solcast's API.

Install

From the directory run the following command:

pip install --user solcast

Tip

for full functionality install pandas: pip install --user solcast pandas

The example notebooks use a variety of optional dependencies to showcase different ways in which the Solcast API may be used. To install these dependencies run

pip install --user solcast[all]

Usage

Warning

To access Solcast data you will need a commercial API key. If you have the API key already, you can use it with this library either as an environment variable called SOLCAST_API_KEY, or you can pass it as an argument api_key when you call one of the library's methods.

Fetching live radiation and weather data:

from solcast import live

res = live.radiation_and_weather(
    latitude=-33.856784,
    longitude=151.215297,
    output_parameters=['air_temp', 'dni', 'ghi']
)

Tip

When testing or developing, you should use solcast.unmetered_locations so that your usage isn't counted against your plan.

Unmetered Locations still require you to signup for a commercial API key (above).

from solcast import forecast
from solcast.unmetered_locations import UNMETERED_LOCATIONS
sydney = UNMETERED_LOCATIONS['Sydney Opera House']

res = forecast.rooftop_pv_power(
    latitude=sydney['latitude'],
    longitude=sydney['longitude'],
    period='PT5M',
    capacity=5,  # 5KW
    tilt=22,  # degrees
    output_parameters='pv_power_rooftop'
)

All response data can be extracted in Python dictionary format.

d = res.to_dict()

If pandas is installed, timeseries responses can be converted to a Pandas DataFrame using the .to_pandas() method.

df = res.to_pandas()

Info

Pandas is not installed by default to keep the environment light.

Available modules are

Module	API Docs
`live`	solcast.live
`historic`	solcast.historic
`forecast`	solcast.forecast
`tmy`	solcast.tmy
`pv_power_sites`	solcast.pv_power_sites
`aggregations`	solcast.aggregations

Contributing & License

Any type of suggestion and code contribution is welcome as PRs and/or Issues. This repository is licensed under MIT (see LICENSE).