Project Agros: analysis of the world’s agricultural output

Description

The repository contains our advanced programming group project, using data on agriculture by Our World in Data, analyzing the world’s agricultural production between the years 1961 and 2019. The goal is to contribute to the green transition.

The project is part of the Advanced Programming Course at Nova SBE. Our main analysis can be found in the showcase Jupyter Notebook. Here, the agricultural evolution of several countries is analyzed with a comparison of the outputs, a gapminder plot, a correlation matrix, area graphs, a choropleth map, and finalized with a prediction of the countries' outputs.

The project is shared work by Group 11:
Julia Stieler - 56040, 56040@novasbe.pt
Hannah Dickescheid - 50178, 50178@novasbe.pt
Carlos Ferrufino - 53276, 53276@novasbe.pt
Eva Zinser - 53100, 53100@novasbe.pt

Data

We are using data on Agricultural total factor productivity (USDA), downloaded from Our World in Data, (datasource can be found here)

The original data was published by the United States Department for Agriculture (USDA) Economic Research Service.

Getting started

1. The environment was saved in the file agrosenv.yml. To learn how to use the environment, follow the documentation found here.

2. Documentation was built using sphinx. To view it, open docs/_build/html/index.html.

3. Our repository is structured like this: In the main directory, you can find our showcase notebook showcase_notebook.ipynb. It contains an analysis of the data and shows our findings. For the project, a class named Agros with several methods was created. All methods used belong to this class. The file agros_class.py containing the class can be found in the python_files directory.

4. Run the download_data method first. It will download the agricultural data from Github and save it to a folder called downloads.

Agros Class

The class is PEP8 compliant, using black and pylint.

Agros has several methods:

1. download_data: downloads agricultural data from Github, saves it to a download folder and creates Pandas DataFrame
2. list_countries: list all the available countries of the dataset
3. correlate_quantity: provides a correlation heatmap of the quality columns
4. area_graph: provides an area graph of the outputs of a selected country or the world
5. compare_output: plots the output columns of selected countries
6. gapminder: provides a scatterplot of fertilizer and output quantity for a selected year
7. choropleth: provided a choropleth plotting the total factor productivity of a selected year
8. predictor: applies an ARIMA prediction for the total factor productivity and plots the data including the prediction

License

GPL-3.0 license

Project Status

Finished

Special Thanks

to Luis and Our World In Data