soybeans.py

from __future__ import print_function

# Sample Gro API client that retrieves data series for Brazil soybeans
# yield from multiple sources.
#
# Usage example: If you have saved GROAPI_TOKEN environment variable,
#
#   python soybeans.py
#
# If you haven't saved the token, you can pass it via cmd line
#
#   python soybeans.py --token ...
#
# Ref https://app.gro-intelligence.com/#/displays/25894

import argparse
import os
import sys

import groclient.lib
from api.client.crop_model import CropModel


def add_brazil_soybeans_yield(model):
    """Adds data series for national level soybeans yield for
    Brazil. There are multiple series from different sources, we add them all."""
    entities = {}
    # Search for item and metric by name, and use the top result
    entities["item_id"] = model.search_for_entity("items", "soybeans")
    entities["metric_id"] = model.search_for_entity("metrics", "yield mass/area")
    # There are many regions with "brazil" in the name, level = 3 is
    # the country
    for region in model.search_and_lookup("regions", "brazil"):
        if region["level"] == 3:
            entities["region_id"] = region["id"]
    data_series_list = model.get_data_series(**entities)
    print("There are {} data series for {}.".format(len(data_series_list), entities))
    for data_series in data_series_list:
        # Look up the name of the source based on the source_id
        source_name = model.lookup("sources", data_series["source_id"]).get("longName")
        print(
            "{}: {} to {}".format(
                source_name, data_series["start_date"], data_series["end_date"]
            )
        )
        model.add_single_data_series(data_series)
    return


def get_brazil_soybeans_weighted_ndvi(model):
    provinces = model.get_provinces("brazil")
    df = model.compute_crop_weighted_series(
        "soybeans",
        "Production Quantity mass",
        "Vegetation NDVI",
        "Vegetation Indices index",
        provinces,
    )
    return df


def main():
    parser = argparse.ArgumentParser(description="Gro api client")
    parser.add_argument("--token", default=os.environ["GROAPI_TOKEN"])
    args = parser.parse_args()

    model = CropModel("api.gro-intelligence.com", os.environ["GROAPI_TOKEN"])
    add_brazil_soybeans_yield(model)
    cwdf = get_brazil_soybeans_weighted_ndvi(model)
    df = model.get_df()
    print("{} data frame with all data series as fetched".format(df.shape))
    print("Columns: {}".format(df.columns))
    print("{} data frame with crop weighted data series".format(cwdf.shape))
    print("Columns: {}".format(cwdf.columns))


if __name__ == "__main__":
    main()