Alluvial plots and analyses for Fire-induced loss of the world’s most biodiverse forests in Latin America 

Address data queries to darmenterasp@unal.edu.co and Bayesian R code related queries to liliana.davalos@stonybrook.edu

1) Place all files in a single folder,

2) alluvial_2step.R makes two-step alluvial plots that require further editing to look as in the paper,

3) alluvial_countries.R makes single step alluvial plots for each country,

4) poisson_overdisperse.R runs MCMCglmm models on the overdispersed Poisson-distributed data and saves them to an RData file needed for plotting coefficients,

5) poisson_plot.R plots the coefficients for the best fit model, requires RData output from step 4.

Note> Explanation of variables 
Land cover
EF, Evergreen forests
DB, Deciduous Broadleaf forests 	
MF, Mixed Forests
WS, Woody Savannas
S, Savannas
G, Grasslands
C, Croplands
Other, other land cover types

Countries
Arg, Argentina
Bol, Bolivia
Bra, Brasil
Chi, Chile
Col, Colombia
Gua, Guatemala
Guy, Guyana
Hon, Honduras
Mex, Mexico
Nic, Nicaragua
Par, Paraguay
Ven, Venezuela


Allowable values are 0-100% in country.csv, proportion_x_country.csv and in all y200*_*f.csv files. 
For country_count.csv and quantity_x_country.csv, values indicate number of pixels