Renewable_Suitability_Predictor

Python Package Index (PyPi): https://pypi.org/project/Renewable-Suitability-Predictor/

This Python package allows the user to predict site suitability for solar or wind energy installations by leveraging machine learning techniques and the NREL Physical Solar Model TMY API (ver 3.2.2)

This package requires that you sign up for an API key at https://developer.nrel.gov/signup/. API rates and limits specified in the NREL documentation apply

To start, install the module

pip install Renewable-Suitability-Predictor

Then import it

import Renwable-Suitability-Predictor as rsp

This module provides the user with two functions

interface_single_prediction(api_key, email, random_state=-1)

interface_multi_prediction(coordinates_array, model_type, email, api_key, random_state=-1)

These functions employ the techniques used in my research paper (https://doi.org/10.58445/rars.682) . They perform 3 main jobs:

1. Gather Typical Meteorological Year (TMY) data from the NREL Physical Solar Model API for each coordinate
2. Transform the time series data to a single dimension
3. Use the Random Forest Classifier (RFC) algorithm to determine suitability

Both functions return a list of two items:

1. A dataframe with the TMY data, prediction, and probability of the prediction for the coordinate(s) the functions are given
2. the metrics from training and testing the RFC model on the dataset used in my research project.

interface_single_prediction uses terminal input from the user to predict the suitability of one location.

interface_multi_prediction is for predicting the suitability of many different locations at once.

Parameter Details

coordinates_array: 2d array-like of shape (n,2). It contains pairs of cooridnates in the order [Latitude, Longitude]. Lat and lon should be given as type float. Example:

coordinates_array = [
    [40.7128, -74.0060],   # New York City, NY
    [34.0549, -118.2426]  # Los Angeles, CA
]

model_type: type string. Either 'wind' or 'solar'. Specifies desired prediction type and whether the solar model or wind model should be used.

api_key: type string. Recieved from the NREL when singing up. Required to interface with the API.

email: type string. Email you used to recieve your API key. Required to interface with the API.

random_state: type int. Used as initial seed for dataframe shuffling, train-test shuffling, and RFC. Used for reproducing function outputs. Default -1 means no set random state. If changed by user, random_state must be in range [0,4294967295]