Replication Data for: The impact of climate change on cacao production in Central America and the Caribbean

These data correspond to the results of the agroclimatic zones of cocoa for the Central America and the Caribbean regions and per country. The data make reference to the ideal zones for cocoa production in three different periods of time (baseline 1970-2000, 2020 - 2049, 2040 - 2069), which allows us to know what would be the impact of climate change for the crop, and therefore, establish a first route for prospective planning in the different countries under study and the region. This analysis was done with a machine learning approach: Random Forest model, which took into account climatic variables, such as precipitation, temperature and evapotranspiration. The model was trained with presence data for construction